Targeting T Cells for the Immune-Modulation of Human Diseases by Regina Lin Department of Immunology Duke University Date:___

Targeting T Cells for the Immune-Modulation of Human Diseases

Regina Lin

Department of Immunology Duke University

Date:______Approved:

______Qi-Jing Li, Supervisor

______You-Wen He, Chair

______Michael S. Krangel

______Xiaoping Zhong

______Xiao-Fan Wang

Dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Immunology in the Graduate School of Duke University

2015

ABSTRACT

Targeting T Cells for the Immune-Modulation of Human Diseases

Regina Lin

Department of Immunology Duke University

Date:______Approved:

______Qi-Jing Li, Supervisor

______You-Wen He, Chair

______Michael S. Krangel

______Xiaoping Zhong

______Xiao-Fan Wang

An abstract of a dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Department of Immunology in the Graduate School of Duke University

2015

Abstract

Dysregulated inflammation underlies the pathogenesis of a myriad of human diseases ranging from cancer to autoimmunity. As coordinators, executers and sentinels of host immunity, T cells represent a compelling target population for immune- modulation. In fact, the antigen-specificity, cytotoxicity and promise of long-lived immune-protection make T cells ideal vehicles for cancer immunotherapy. Interventions for autoimmune disorders, on the other hand, aim to dampen T cell-mediated inflammation and promote their regulatory functions. Although significant strides have been made in targeting T cells for immune-modulation, current approaches remain less than ideal and leave room for improvement. In this dissertation, I seek to improve on current T cell-targeted immunotherapies, by identifying and preclinically characterizing their mechanisms of action and in vivo therapeutic efficacy.

CD8+ cytotoxic T cells have potent antitumor activity and therefore are leading candidates for use in cancer immunotherapy. The application of CD8+ T cells for clinical use has been limited by the susceptibility of ex vivo-expanded CD8+ T cells to become dysfunctional in response to immunosuppressive microenvironments. To enhance the efficacy of adoptive cell transfer therapy (ACT), we established a novel microRNA- targeting approach that augments CTL cytotoxicity and preserves immunocompetence.

Specifically, we screened for miRNAs that modulate cytotoxicity and identified miR-23a as a strong functional repressor of the transcription factor Blimp-1, which promotes CTL cytotoxicity and effector cell differentiation. In a cohort of advanced lung cancer patients,

miR-23a was upregulated in tumor-infiltrating CD8+ T cells, and its expression correlated with impaired antitumor potential of patient CD8+ T cells. We determined that tumor-derived TGF-β directly suppresses CD8+ T cell immune function by elevating miR-23a and downregulating Blimp-1. Functional blockade of miR-23a in human CD8+

T cells enhanced granzyme B expression; and in mice with established tumors, immunotherapy with just a small number of tumor-specific CD8+ T cells in which miR-

23a was inhibited robustly hindered tumor progression. Together, our findings provide a miRNA-based strategy that subverts the immunosuppression of CD8+ T cells that is often observed during adoptive cell transfer tumor immunotherapy and identify a TGFβ- mediated tumor immune-evasion pathway.

Having established that miR-23a-inhibition can enhance the quality and functional-resilience of anti-tumor CD8+ T cells, especially within the immune- suppressive tumor microenvironment, we went on to interrogate the translational applicability of this strategy in the context of chimeric antigen receptor (CAR)-modified

CD8+ T cells. Although CAR T cells hold immense promise for ACT, CAR T cells are not completely curative due to their in vivo functional suppression by immune barriers ‒ such as TGFβ ‒ within the tumor microenvironment. Since TGFβ poses a substantial immune barrier in the tumor microenvironment, we sought to investigate whether inhibiting miR-23a in CAR T cells can confer immune-competence to afford enhanced tumor clearance. To this end, we retrovirally transduced wildtype and miR-23a-deficient

CD8+ T cells with the EGFRvIII-CAR, which targets the PepvIII tumor-specific epitope

expressed by glioblastomas (GBM). Our in vitro studies demonstrated that while wildtype EGFRvIII-CAR T cells were vulnerable to functional suppression by TGFβ, miR-23a abrogation rendered EGFRvIII-CAR T cells immune-resistant to TGFβ.

Rigorous preclinical studies are currently underway to evaluate the efficacy of miR-23a- deficient EGFRvIII-CAR T cells for GBM immunotherapy.

Lastly, we explored novel immune-suppressive therapies by the biological characterization of pharmacological agents that could target T cells. Although immune- suppressive drugs are classical therapies for a wide range of autoimmune diseases, they are accompanied by severe adverse effects. This motivated our search for novel immune- suppressive agents that are efficacious and lack undesirable side effects. To this end, we explored the potential utility of subglutinol A, a natural product isolated from the endophytic fungus Fusarium subglutinans. We showed that subglutinol A exerts multimodal immune-suppressive effects on activated T cells in vitro: subglutinol A effectively blocked T cell proliferation and survival, while profoundly inhibiting pro- inflammatory IFNγ and IL-17 production by fully-differentiated effector Th1 and Th17 cells. Our data further revealed that subglutinol A might exert its anti-inflammatory effects by exacerbating mitochondrial damage in T cells, but not in innate immune cells or fibroblasts. Additionally, we demonstrated that subglutinol A significantly reduced lymphocytic infiltration into the footpad and ameliorated footpad swelling in the mouse model of Th1-driven delayed-type hypersensitivity. These results suggest the potential of subglutinol A as a novel therapeutic for inflammatory diseases.

Dedication

To Mom, Dad, Jose and Bacon.

vii

Contents

Abstract ...... iv

List of Tables ...... xii

List of Figures ...... xiii

Acknowledgements ...... xvii

1. Introduction ...... 1

1.1 Cancer immunotherapy: An overview ...... 2

1.1.1 Hurdles to successful cancer immunotherapy ...... 4

1.1.2 Cytokine-based approaches...... 10

1.1.3 Antibody-based approaches ...... 15

1.1.4 Cell-based approaches...... 18

1.2 CD8+ T cells in cancer immunotherapy ...... 23

1.2.1 Immune-protection conferred by effector and memory CD8+ T cells ...... 23

1.2.2 Anti-tumor effector functions of CD8+ T cells ...... 26

1.2.3 Transcriptional regulation of CD8+ T cell effector functions and memory generation ...... 31

1.3 Chimeric antigen receptors (CARs): Novel gene therapy tools for ACT ...... 36

1.3.1 Utility of CAR T cell-based ACT ...... 37

1.3.2 Current challenges facing CAR T cell-based ACT ...... 39

1.4 MicroRNAs as targets for T cell immune-modulation ...... 41

1.4.1 Roles of microRNAs in CD8+ T cell fate and function ...... 41

1.4.2 Tools for targeting microRNAs ...... 43

viii

1.5 Targeting auto-reactive effector T cells for the treatment of inflammatory diseases ...... 46

1.5.1 The unmet need for tolerable T cell-specific immune-suppressants ...... 46

2. Materials and Methods ...... 49

2.1 Mice ...... 49

2.2 Cell culture ...... 49

2.3 miRNA expression profiling and miRNA qPCR ...... 50

2.4 Target prediction and luciferase reporter assays ...... 51

2.5 mRNA and pri-miRNA qPCR ...... 52

2.6 Lymphocyte isolation and miR-23a quantification from lung cancer patients ..... 52

2.7 miR-23a decoy construct design ...... 53

2.8 Retroviral transduction...... 54

2.8.1 miR-23a overexpression and decoy vectors ...... 54

2.8.2 EGFRvIII-CAR vector ...... 54

2.9 Western blot ...... 55

2.10 In vitro cytotoxicity assays ...... 55

2.11 CD4+ T cell differentiation assays ...... 56

2.12 T cell restimulation ...... 56

2.13 Intracellular staining and flow cytometry ...... 57

2.14 Proliferation and cell viability assays ...... 58

2.15 Labeling and functional analysis of mitochondria ...... 58

2.16 Delayed-type hypersensitivity induction and treatment ...... 59

2.17 In vivo tumor models ...... 60

2.17.1 Subcutaneous tumor models ...... 60

2.17.2 Intracerebral glioblastoma model...... 61

2.18 T cell repertoire library construction ...... 61

2.19 Statistical analysis ...... 62

3. Mechanistic significance of targeting miR-23a in CD8+ T cells for ACT ...... 63

3.1 Introduction ...... 63

3.2 Results ...... 66

3.2.1 Identification of miR-23a expression as a negative correlate of cytotoxicity of effector CTLs ...... 66

3.2.2 Forced miR-23a expression compromises anti-tumor CTL effector responses in vivo ...... 80

3.2.3 Functional blockade of miR-23a in CTLs augments their anti-tumor function in vitro ...... 83

3.2.4 miR-23a blunts CTL effector responses by targeting Blimp-1 ...... 89

3.2.5 In effector CTLs, TCR activation and TGFβ signaling differentially regulate miR-23a expression ...... 94

3.2.6 TCR and TGFβ signaling converge on cMyc to differentially modulate miR- 23a expression in effector CTLs ...... 102

3.2.7 Neutralizing miR-23a in CTLs mitigates TGFβ-induced immunosuppression ...... 107

3.2.8 miR-23a expression correlates inversely with anti-tumor potential of mouse and human tumor-infiltrating CD8+ T cells ...... 110

3.2.9 Adoptive transfer of miR-23a-inhibited CTLs robustly retard tumor progression ...... 116

3.3 Discussion ...... 123

4. Extending the translational utility of ACT through CD8+ CAR T cells ...... 127

4.1 Introduction ...... 127

4.2. Results ...... 130

4.2.1 Targeting miR-23a in EGFRvIII-CAR T cells confers resistance to TGFβ- induced suppression ...... 130

4.2.2 EGFRvIII-CAR therapy drives the mobilization and convergence of endogenous intratumoral T cell repertoires ...... 136

4.3 Discussion ...... 147

5. Biological evaluation of subglutinol A as a novel immunosuppressive agent for the intervention of T cell-mediated inflammatory diseases ...... 151

5.1 Introduction ...... 151

5.2 Results ...... 154

5.2.1 Subglutinol A inhibits the expansion of activated T cells ...... 154

5.2.2 Subglutinol A abrogates IL-2 production by activated CD4+ T cells ...... 157

5.2.3 Subglutinol A abolishes pro-inflammatory cytokine production by effector Th1 and Th17 cells ...... 160

5.2.4 Subglutinol A preferentially exacerbates mitochondrial depolarization in effector Th1 and Th17 cells ...... 163

5.2.5 Subglutinol A treatment suppresses Th1-driven delayed-type hypersensitivity (DTH) reactions in vivo ...... 168

5.3 Discussion ...... 171

6. General discussion and future perspectives ...... 172

References ...... 182

Biography ...... 231

List of Tables

Table 1: 18 miRNAs significantly differentially expressed in DC- and B cell primed CTLs, as determined by the paired t-test with significance level set at 0.05...... 74

xii

List of Figures

Figure 1: Phenotype of splenic B cells and LPS-matured BMDCs (mDCs) used for naïve CTL priming in vitro...... 69

Figure 2: Priming by splenic B cells and mDCs generate CTLs with different cytotoxic profiles...... 70

Figure 3: Exogeneous IL-15 and IL-21 partially rescued granzyme B and activation- induced cell death, but not cytotoxicity of B cell-primed CTLs...... 71

Figure 4: Volcano plot of miRNA expression profiling...... 72

Figure 5: Heatmap of miRNAs differentially-expressed by DC- and B cell-primed CTLs...... 73

Figure 6: miR-23a inhibited both granzyme B and T-bet expression in CTLs in vitro. ... 75

Figure 7: miR-23b, the paralog of miR-23a, did not affect Granzyme B and T-bet expression in CTLs...... 76

Figure 8: Validation of differential miR-23a expression following in vitro priming...... 77

Figure 9: miR-23a does not affect proliferation or AICD of activated CTLs...... 78

Figure 10: Quantification of miR-23a expression levels in activated murine CTLs...... 79

Figure 11: Forced expression of miR-23a in tumor-specific CTLs impairs tumor retardation, but does not alter their intratumoral accumulation...... 81

Figure 12: Forced expression of miR-23a in tumor-specific CTLs inhibits their anti-tumor effector responses in vivo...... 82

Figure 13: LNA-mediated functional blockade of miR-23a enhances CTL effector responses ex vivo...... 85

Figure 14: Inhibiting miRNAs unrelated to miR-23a do not enhance CTL effector function...... 86

Figure 15: The miR-23a decoy retroviral expression vector...... 87

Figure 16: Decoy-mediated functional blockade of miR-23a enhances CTL effector responses ex vivo...... 88

xiii

Figure 17: miR-23a does not repress the expression of glutaminase in effector CTLs. ... 91

Figure 18: Computationally-predicted miR-23a targets relevant to CTL effector function...... 92

Figure 19: Blimp-1 is a direct target of miR-23a in CTLs...... 93

Figure 20: TCR activation, but not stimulation strength, suppresses miR-23a expression in CTLs...... 97

Figure 21: Co-receptor signaling through CD28, CD40L and PD-1 do not affect miR-23a expression in CTLs...... 98

Figure 22: Notch signaling does not regulate miR-23a expression in CTLs...... 99

Figure 23: Cytokines that do not alter miR-23a expression in CTLs...... 100

Figure 24: TGFβ promotes miR-23a expression in CTLs...... 101

Figure 25: TCR activation-induced cMyc represses pri-miR-23a transcription in effector CTLs...... 104

Figure 26: TGFβ suppresses cMyc activity in effector CTLs...... 105

Figure 27: cMyc is the major repressor of pri-miR-23a transcription in primed CTLs. 106

Figure 28: Neutralizing miR-23a in CTLs mitigates TGFβ-induced immunosuppression...... 109

Figure 29: miR-23a is up-regulated in mouse and human CD8+ TILs...... 112

Figure 30: miR-23a levels correlate inversely with Blimp-1 expression in human CD8+ T cells...... 113

Figure 31: miR-23a expression correlates inversely with anti-tumor effector functions of human CD8+ T cells...... 114

Figure 32: miR-23a blockade boosts the effector function of human CD8+ T cells...... 115

Figure 33:Adoptive T cell transfer therapy with miR-23a-inhibited CTLs dramatically retarded the progression of established tumors in a mouse model of melanoma...... 118

xiv

Figure 34: Adoptive T cell transfer therapy with miR-23a-inhibited CTLs dramatically retarded the progression of established tumors in a mouse model of subcutaneous Lewis lung cancer...... 119

Figure 35: Intratumoral persistence of miR-23a-inhibited CTLs was unaffected...... 120

Figure 36: miR-23a inhibition enhanced the anti-tumor effector function of tumor- specific CTLs in vivo...... 121

Figure 37: Inhibiting granzyme B in pMel-1 CTLs abrogates the therapeutic advantage conferred by the miR-23a Decoy...... 122

Figure 38: Model of CTL immune-modulation by targeting miR-23a...... 126

Figure 39: miR-23a deletion efficiency in pMel-1 ER-Cre+ miR-23af/f CD8+ T cells. 133

Figure 40: Transduction of pMel-1 CD8+ T cells with the 3 rd generation EGFRvIII-CAR retroviral vector...... 134

Figure 41: miR-23a deletion in EGFRvIII-CAR T cells confers resistance to TGFβ- induced functional suppression...... 135

Figure 42: EGFRvIII expression on the KLucvIII murine GBM cell line...... 140

Figure 43: Comparable tumor burdens at point of tissue sample collection...... 142

Figure 44: Distribution and clonal frequencies of IC-delivered CAR T cells and endogenous T cells...... 144

Figure 45: Tumor antigens drive the local accumulation of shared tumor-reactive T cell clones...... 145

Figure 46: vIII-CAR WT therapy drives the convergence of inter-individual intratumoral TCR repertoires...... 146

Figure 47: Structure of subglutinols A and B...... 153

Figure 48: Subglutinol A blocks antigen-induced T cell proliferation and induces massive apoptosis...... 156

Figure 49: Subglutinol A abrogates antigen-induced IL-2 production by activated CD4+ T cells...... 159

Figure 50: Subglutinol A abolishes inflammatory cytokine production in fully differentiated Th1 and Th17 cells...... 162

Figure 51: Subglutinol A exacerbates mitochondrial depolarization in fully differentiated Th1 and Th17 cells in vitro...... 166

Figure 52: Subglutinol A does not affect mitochondrial mass or polarization in dendritic cells and stromal cells in vitro...... 167

Figure 53: Subglutinol A treatment suppresses antigen-induced DTH responses in vivo...... 170

xvi

Acknowledgements

I would like to acknowledge the many people who have helped and supported me over the past 6 years.

First, I want to thank my PhD advisor, Dr. Qi-Jing Li., who has been an excellent mentor. His guidance, encouragement and enthusiasm have been instrumental to my scientific and intellectual development. Dr. Li’s doors are always open for scientific discussion or otherwise, making him a great pleasure to work with. I am almost certain that graduate school would have been less colorful without an exuberant mentor like Dr.

Li.

I would also like to thank the past and present members of the Li Lab, especially

Shan, Chaoran, Jose, Erik, Siqi, Elizabeth and Laura, for their scientific input, technical support, and friendly banter. Everyone has helped forge the conducive and collaborative working environment that makes the Li Lab.

I want to express my gratitude to my thesis committee – Dr. You-Wen He, Dr.

Mike Krangel, Dr. Xiaoping Zhong and Dr. Xiao-Fan Wang. Their insights and advice over the years have helped strengthen my research and kept me on the right track.

I am also grateful for the scientific input, technical support and reagents generously extended to me by other laboratories at Duke Immunology, as well as my collaborators. I want to thank the members of the Flow Cytometry Core, particularly

Mike, Nancy and Lynn, for their assistance and patience with cell sorting. I would like to extend my gratitude to Dr. Feng Feng at Boston University, who kindly helped analyze

xvii

the miRNA expression profiling data. I also really appreciate the help provided by the

Sampson Lab in Duke Neurosurgery, especially Dr. John Sampson, Dr. Luis-Sanchez

Perez and Carter Suryadevara, in moving along the CAR project. I would also like to specially thank Dr. Jiyong Hong from the Department of Chemistry, who dedicated his time, effort and commitment in bringing the Subglutinol A project to fruition.

In addition, I want to thank my classmates at Duke Immunology, particularly

Young Joo, Kathleen, Sarah and Guang. They make the bad times tolerable and the good times so much more fun. I am fortunate to have come to know friends like these.

I am grateful for my family in Singapore – Mom, Dad, my brother Ray and my sister-in-law Xinrong. Mom and Dad have always given me unconditional love, support, understanding and encouragement, no matter the distance. I also want to thank Ray and

Xinrong for always taking such great care of my parents; my mind can be at ease knowing that my parents are in good hands.

Finally, I want to thank my closest family here in the US – Jose and Bacon. Jose has been and will continue to be my main pillar of support. During the roughest of times, he has always stood by me, motivated me and never stopped believing in my abilities.

During the good times, he never forgets to celebrate even the littlest of my victories. As for Bacon, he provides a vital, on-demand source of cuteness and comic relief. I would not have been able to get this far without them.

xviii

1. Introduction

Dysregulated immune responses – either hypo-responsiveness or hyper- responsiveness – underlie a plethora of chronic human diseases ranging from cancer to autoimmunity. In the case of cancer, the breakdown of host immune defense mechanisms results in the defective eradication and consequent outgrowth of malignant cells. On the flipside, aberrant and excessive immune activation culminates in inflammatory disorders.

That the immune system plays an etiological role in disease pathogenesis has inspired the notion of immune-modulation for therapeutic intervention.

Among the various immune cell types, the population most commonly implicated in immune-dysregulation is perhaps T cells. Given their diverse effector subsets (e.g.

Th1, Th2, Th17 and Treg) and memory differentiation statuses (e.g. central, effector or resident memory), T cells are uniquely positioned as key orchestrators and executers in virtually all aspects of both short- and long-term immune responses. Critical regulatory factors and their downstream signaling pathways controlling the magnitude of T cell effector responses, as well as the various stages of memory differentiation have now been identified, thereby offering the prospect of manipulating T cell function for immune- modulation. Taken together, their multi-faceted involvement in disease pathogenesis, coupled to their functional plasticity, poise T cells as a compelling population for therapeutic targeting. In this chapter, we will review the advantages and limitations of current immune-modulatory approaches for cancer and autoimmunity, with a focus on the recent efforts invested towards T cell-based immunotherapies.

1.1 Cancer immunotherapy: An overview

Cancer immunotherapy aims to harness the effector and surveillance functions of the immune system, in order to achieve the short-term goal of tumor regression and the long-term goal of preventing tumor recurrence. The concept of cancer immunotherapy stems from the immunogenicity of tumor cells: that tumor cells differ antigenically from normal cells enables them to trigger immune responses. Indeed, it has been demonstrated in melanoma patients that melanoma expressing the tumor-specific antigen, MAGE-1, can induce endogenous MAGE-1-specific CD8+ T cells, which are capable of eliciting antigen-specific melanoma destruction (van der Bruggen et al., 1991).

Virtually all branches of innate and adaptive immunity participate in the successful recognition, elimination and long-term surveillance of transformed cells.

Necrotic tumor cells release damage-associated molecular patterns (DAMPs) (e.g. dsDNA) and alarmins (e.g. the TLR4 ligand, HMGB1) recognized by tissue-resident macrophages and dendritic cells (DCs) (Apetoh et al., 2007; Deng et al., 2014; Woo et al., 2014). These activate STING-dependent and TLR-mediated signaling in macrophages and DCs to trigger their functional maturation, as characterized by the up-regulation of surface major histocompatibility complexes (MHC) and co-stimulatory molecules (e.g.

CD80 and CD86), as well as the production of pro-inflammatory cytokines (e.g. Type 1 interferons, IL-12 and TNFα) (Deng et al., 2014; Fuertes et al., 2013; Gallucci et al.,

1999; Sauter et al., 2000; Woo et al., 2014). Tumor cell debris picked up by the scavenger receptors on DCs and macrophages are phagocytosed and degraded in lysosomes, generating tumor-derived peptides that are loaded onto MHCII for

presentation to CD4+ T cells, and onto MHC I for cross-presentation to CD8+ T cells

(Bedoui et al., 2009; Gallucci et al., 1999). Collectively, innate immune sensing of tumors promotes the activation of adaptive immunity. Tumor-reactive T cells recognizing their cognate antigen in the context of MHC and co-stimulatory signals become activated and acquire effector functions, which include DC and macrophage licensing (e.g. through

CD40:CD40L ligation) (Ahmed et al., 2012) to further perpetuate innate immune responses, Type-1 cytokine production (e.g. IFNγ and IL-2), as well as the induction of the cytolytic machinery (e.g. granzyme B and perforin) (Elgueta et al., 2009; Hivroz et al., 2012). Upon re-encountering these antigens presented on tumor MHCI molecules, tumor-reactive CD8+ T cells can mediate direct tumor cell lysis. Even tumor cells that evade T cell recognition through MHCI down-regulation can be eliminated, as this

“missing self” triggers NK cell-mediated killing (Carlyle et al., 2004). In addition, NK cell-mediated tumor destruction is also promoted by the expression of NK cell-activating ligands (e.g. the NKG2D ligands MICA/B and ULBPs) on tumor cells (Baragano

Raneros et al., 2014; Bauer et al., 1999; Guerra et al., 2008). Anti-tumor antibodies produced by B cells opsonize tumor cells, facilitating Fc receptor-mediated phagocytosis, antibody-dependent cytotoxicity and complement-mediated lysis of tumor cells (Biburger et al., 2014). Memory T and B cells that develop not only ensure long-lasting immune- surveillance, but also are able to respond rapidly in the event of tumor recurrence. Under steady state, these safeguards are constantly at play to survey for and eradicate aberrant, pre-malignant cells, preventing the establishment and development of tumors.

This initial immune-activation, however, imposes a selection pressure on tumor cells. Through the process of immune-editing, the outgrowth of less immunogenic or immune-evasive tumor variants that escape immune clearance is favored (Quezada et al.,

2011). Therefore, at the point of tumor establishment and clinical diagnosis, anti-tumor immune responses have usually become ineffective at keeping tumor progression in check, either due to immune-escape of antigen-loss tumor variants, or due to tumor- induced immune-suppression (Crespo et al., 2013; Flavell et al., 2010; Garrido et al.,

2010). Reversing this immune-tolerance to reactivate and reinforce robust anti-tumor immune responses, therefore, lie at the heart of cancer immunotherapy.

Fortunately, the diverse immune subsets involved in anti-tumor defense, coupled to the complex interactions between immune and tumor cells, provide therapeutic targets for intervention at multiple levels. In the design of efficacious immunotherapeutic strategies, it is important to bear in mind the antigenic properties and unique immune- suppressive milieu of tumors. Key parameters for consideration, as well as immunotherapeutic strategies aimed at overcoming these hurdles, are discussed below.

1.1.1 Hurdles to successful cancer immunotherapy

A primary challenge facing effective cancer immunotherapy is target antigen selection. Cancer immunotherapy relies on immune recognition of specific tumor antigens, which in turn sets off a cascade to amplify anti-tumor effector responses. Safety is a requisite in selecting a target antigen for clinical development: ideally, the optimal target antigen should direct immune responses to specifically eliminate tumor cells, without causing damage to normal host cells. The process of tumorigenesis provides 4

candidate antigens for immune-targeting: genetically-unstable malignant cells acquire genetic aberrations resulting in (i) tumor-associated antigens (TAAs) that are shared with normal cells, but whose expression levels (e.g. MART-1 in melanoma) (Kawakami et al.,

1994) or post-translational modifications are altered in tumor cells (e.g. Muc1 in breast and ovarian cancers) (Girling et al., 1989), and (ii) tumor-specific neo-antigens (TSAs) that are uniquely expressed by tumor cells, including truncated proteins (e.g. EGFRvIII in glioblastoma) (Sugawa et al., 1990; Wong et al., 1992; Yamazaki et al., 1990).

Particularly in less immunogenic tumors, a large majority of genetic aberrations give rise to TAAs, rather than TSAs (Vogelstein et al., 2013). One the one hand, unmutated TAAs serve as more cost-effective therapeutic targets, since they are shared by most patients.

On the other hand, this also makes it challenging to identify and select TAAs or epitopes in TAAs that can specifically target tumors, while sparing normal tissues from bystander damage (Overwijk et al., 2003; Palmer et al., 2008). To circumvent these “off-tumor” effects, the field of ACT is geared increasingly towards the development of TSA- targeting personalized immunotherapies: TSAs unique to malignant cells can now be identified by cancer whole-exome sequencing, enabling the ex vivo selection and expansion of TSA-specific autologous T cells (Linnemann et al., 2015; Tran et al., 2014;

Yadav et al., 2014). Albeit therapeutically efficacious, the TSAs identified thus far are unique to individual patients (Linnemann et al., 2015), making these highly personalized approaches extremely labor-intensive and cost-prohibitive.

Another obstacle is the difficulty in isolating therapeutically sufficient numbers of anti-tumor immune effector cells, due primarily to two reasons. First, the immune

system’s remarkable ability to discriminate between self and non-self poses a conundrum, as mature immune cells have little reactivity for tumor cells that are perceived as self- derived. For T cells in particular, clones bearing high-avidity self-reactive T cell receptors (TCRs) are either eliminated during thymic development, mature into regulatory T cells, or exit the thymus but are rendered anergic in the periphery. These mechanisms of immune-tolerance sculpt the host’s TCR repertoire, such that mature, functional conventional T cells in the periphery – the main executers of tumor sensing, clearance and long-term surveillance – are largely unable to recognize self-derived tumor cells and TAAs (Tran et al., 2014; Yadav et al., 2014). Second, as the incidence of cancer increases with age, TCR diversity wanes. After reaching its peak function around puberty, the thymus involutes, dramatically reducing the thymic output of naïve T cells in the adult (Naylor et al., 2005). This age-dependent decline in naïve T cell generation translates into a narrowed peripheral TCR repertoire (Kohler et al., 2005; Rudd et al.,

2011), further limiting the spectrum and responsiveness of endogenous T cells bearing tumor-reactivity (Yager et al., 2008). Therefore, although functional endogenous T cells specific for TSAs can be found in cancer patients, they are present at low frequencies

(Linnemann et al., 2015; Tran et al., 2014), making it challenging to isolate them for autologous reinfusion.

A third hurdle is the presence of immunosuppressive barriers within the tumor microenvironment, which are co-opted by tumors to evade the host immune system

(Fourcade et al., 2010; Zou, 2005). These may include soluble factors, regulatory immune cells and immune checkpoints.

Soluble factors include cytokines, such as TGFβ, IL-10 and IL-6, as well as the metabolic enzyme indoleamine 2,3-dioxygenase (IDO). Of these, TGFβ poses the greatest immune barrier. TGFβ is secreted and up-regulated by a wide variety of tumors, including melanoma and lung cancer (Bennicelli and Guerry, 1993; De Jaeger et al.,

2004; Derynck et al., 1985). Elevated serum TGFβ has been found to correlate with increased rates of tumor recurrence and metastasis, along with decreased overall survival

(Krasagakis et al., 1998; Zhao et al., 2010). As a pleiotropic cytokine, TGFβ promotes tumor metastasis, while attenuating both innate and adaptive anti-tumor immune responses. TGFβ interferes with tumor antigen-presentation by DCs, either by inhibiting their migration to tumor draining lymph nodes (Weber et al., 2005), or by inducing DC apoptosis (Ito et al., 2006). Moreover, NK cell-mediated tumor killing can be impaired by

TGFβ, due to down-regulation of the killer activating receptors, NKG2D and NKp30

(Castriconi et al., 2003). TGFβ suppresses IL-2 production by effector T cells (Brabletz et al., 1993) and induces cell cycle arrest (McKarns et al., 2004; Ruegemer et al., 1990;

Stephen et al., 2014; Wolfraim et al., 2004), two consequences that cooperatively inhibit

T cell expansion within the tumor. Besides exerting anti-proliferative effects on T cells,

TGFβ also impacts the functional differentiation of both CD4+ and CD8+ T cells. The

TGFβ-enriched tumor microenvironment dampens TCR signaling-induced expression of the Th1 transcription factor T-bet (Chen et al., 2003; Gorelik et al., 2002), and favors

Smad3-induced expression of the FoxP3 transcription factor (Tone et al., 2008). This effectively polarizes tumor-reactive CD4+ T cells from Th1 effectors towards the inducible regulatory T cell (iTreg) lineage (Gu et al., 2012; Liu et al., 2007). In both

naïve and full-fledged effector CD8+ T cells, TGFβ-induced Smad signaling represses their expression of key cytotoxic mediators, including granzyme B and IFN γ, resulting in

CD8+ T cell dysfunction and impaired tumor rejection (Ahmadzadeh and Rosenberg,

2005; Stephen et al., 2014; Thomas and Massague, 2005).

Finally, immune checkpoints can be exploited by tumors to subvert immune clearance. Immune checkpoints, such as CTLA-4, PD-1 and LAG-3, are built-in safeguards that curtail the response of effector cells to prevent immune hyperactivation and maintain self-tolerance. CTLA-4, PD-1 and LAG-3 are co-inhibitory receptors expressed on activated T cells: while CTLA-4 functions at the early stages of T cell activation and priming in lymphoid organs, PD-1 and LAG-3 predominantly acts in peripheral sites to limit normal tissue damage by effector cells (Keir et al., 2006;

Waterhouse et al., 1995). Although all three molecules are not expressed on naïve T cells, their surface expression is rapidly induced upon T cell activation (Agata et al., 1996;

Annunziato et al., 1996). However, while CTLA-4 expression is restricted to activated T cells, PD-1 and LAG-3 are also expressed by a broad variety of activated non-T cells, including B cells and NK cells (Agata et al., 1996; Huard et al., 1994; Terme et al.,

2011). Like CD28, CTLA-4 binds to CD80 and CD86 on APCs, albeit with a much higher affinity, resulting in preferential binding to CTLA-4 (Linsley et al., 1994). CTLA-

4 ligation not only transduces inhibitory signals into the activated T cell (Riley et al.,

2002; Schneider et al., 2002), but also deprives it of co-stimulatory signals through the sequestration or physical removal of CD80 and CD86 from the APC surface (Qureshi et

al., 2011). Therefore, the engagement of CTLA-4 prevents optimal priming of anti-tumor

T cells in the draining lymph nodes.

Under normal physiological conditions, ligation of PD-1 on effector cells by either PD-L1 or PD-L2 on mature APCs reins in the immune response to prevent autoimmunity (Yamazaki et al., 2002). However, this homeostatic pathway is frequently co-opted by tumors for immune-evasion. PD-1 ligands have been found to be up- regulated in as many as 50-70% of human cancers, and their expression correlates with poorer prognosis (Hamanishi et al., 2007; Nomi et al., 2007; Ohigashi et al., 2005;

Topalian et al., 2012). Moreover, due to chronic antigen exposure, tumor-infiltrating T cells in human patients express high levels of PD-1 (Tumeh et al., 2014).

Mechanistically, PD-1 ligation on T cells triggers the recruitment of the SHP2 phosphatase, which dephosphorylates the TCR signaling complex and attenuates T cell activation (Yokosuka et al., 2012). Interaction between PD-1 on CTLs and PD-L1 on tumor or stromal cells potently curtails intratumoral CTL expansion and cytotoxic potency, leading to impaired tumor rejection (Fourcade J, 2010; Iwamura K, 2012;

Tumeh et al., 2014; Woo SR, 2012).

LAG-3, a structural homolog of CD4, binds MHCII with remarkable affinity

-8 -4 (LAG-3 K d ~ 10 versus CD4 K d ~10 ) (Huard et al., 1995; Triebel et al., 1990). LAG-3 exerts diverse immune-modulatory effects on both effector and regulatory immune cells.

On the one hand, the engagement of MHCII on innate APCs by LAG-3 induces their functional maturation, marked by enhanced co-stimulatory molecule expression and pro- inflammatory cytokine secretion, which collectively drive Type 1 adaptive anti-tumor

responses (Andreae et al., 2002). On the other hand, the ligation of LAG-3 on effector T cells and NK cells transmits an inhibitory signal that restricts their proliferation and cytotoxic functions (Grosso et al., 2007; Macon-Lemaitre and Triebel, 2005).

Furthermore, triggering LAG-3 on activated nTregs and iTregs augments their immune- suppressive activity (Huang et al., 2004). LAG-3-expressing TILs have been found across multiple human tumors (e.g. renal cell carcinoma, melanoma and ovarian cancer) and their expression on tumor-reactive TILs correlates with impaired effector functions

(Demeure et al., 2001; Matsuzaki et al., 2010), underscoring the role of LAG-3 in hampering immune-mediated tumor rejection.

1.1.2 Cytokine-based approaches

3 signals are required to instruct an appropriate T cell response: 1) TCR signaling triggered by antigen recognition, 2) co-stimulatory signals and 3) cytokines that orchestrate T cell differentiation. To support the first two signals and skew endogenous immune responses towards an anti-tumor phenotype, therapeutic approaches involving the systemic administration of immune-potentiating cytokines have been developed.

Particular attention has been given to cytokines that signal through the common gamma- chain (γc), primarily due to their ability to support NK cell and T cell anti-tumor responses. In clinical trials with IL-7, IL-21 and IL-15, all three agents demonstrated safety and partial efficacy. However, the efficacy of cytokine therapy is still contingent on the availability of Signals 1 and 2. Therefore, cytokine monotherapy is less efficacious than when co-administered as an adjuvant with either tumor vaccines or checkpoint blockade therapies, or as a complement to ACT. 10

Systemic IL-7 administration provides a global boost to T cell responses. In preclinical studies, IL-7 treatment decreased the expression of Cbl-b, a negative regulator of TCR signaling expression in T cells, thereby lowering the threshold for T cell activation (Pellegrini et al., 2009); this enabled the mobilization of T cell responses against weak or subdominant antigens. In addition, IL-7 reduces the sensitivity of T cells to TGFβ-mediated suppression, by down-regulating surface TGFβRII expression and decreasing the phosphorylation of Smad2 and Smad3 (Pellegrini et al., 2009). Phase 1 clinical trials demonstrated that IL-7 was generally well-tolerated (Sportes et al., 2010).

When administered systemically into patients, IL-7 preferentially enhanced the polyclonal proliferation of recent thymic emigrants, naive conventional T cells and central memory T cells, but not terminally-differentiated effector T cells or Tregs

(Sportes et al., 2010; Sportes et al., 2008). The non-potentiating effects of IL-7 on Tregs stems from the fact that Tregs do not express IL-7Rα, but instead constitutively express the high affinity IL-2Rα (CD25) and are therefore highly-dependent on IL-2 (Powell et al., 2007; Seddiki et al., 2006). This makes IL-7 a superior agent for cytokine immunotherapy than IL-2: in side-by-side comparisons with IL-2, IL-7 was more therapeutically efficacious as the selective augmentation of conventional T cells could more effectively overcome Treg-mediated suppression (Perna et al., 2014). Moreover, by enhancing cycling of naïve conventional T cells, IL-7 treatment increased the diversity of patients’ TCR repertoires (Sportes et al., 2008). Taken together, IL-7 therapy rejuvenated patients’ immune systems, such that their T cell compartments became enriched in non- terminally differentiated T cell subsets capable of a broader spectrum of antigen

recognition. However, IL-7 therapy is not without shortfalls. Although IL-7 provides an overall boost to T cell numbers, it does not enhance their cytolytic potency on a per-cell basis (Markley and Sadelain, 2010). Additionally, the therapeutic window for IL-7 may be narrow due to the down-regulation of IL-7Rα in anti-tumor effector T cells upon chronic IL-7 exposure or upon antigen stimulation (Alves et al., 2008). These may explain the findings from preclinical studies, in which incorporating IL-7 into ACT only modestly improved tumor clearance, with a therapeutic effect that quickly saturated at low to intermediate IL-7 doses (Klebanoff et al., 2011).

IL-21 is another immune-potentiating cytokine that has entered clinical testing.

Phase I and II clinical trials of IL-21 in melanoma and renal cell carcinoma have revealed that a biologically-efficacious dose of IL-21 elicits multiple adverse events in some patients, ranging from mild flu-like symptoms to more severe lymphopenia and hepatotoxicity (Davis et al., 2009; Davis et al., 2007; Thompson et al., 2008). IL-21 supports the survival, proliferation and granzyme B expression in CTLs (Novy et al.,

2011; Zeng et al., 2005). Furthermore, ACT immunotherapy in mouse tumor models demonstrated that co-administration of systemic IL-21 promoted the differentiation of adoptively-transferred tumor-specific cells into central memory CD8+ T cells, which conferred superior anti-tumor protection in vivo (Hinrichs et al., 2008). Preclinical trials and Phase I trials revealed that systemic IL-21 administration not only promoted the activation, expansion and cytolytic potency of both NK cells and CTLs (Frederiksen et al., 2008; Steele et al., 2012), but also enhanced memory T cell generation (Markley and

Sadelain, 2010). However, these measures of immune-activation did not translate into

clinical efficacy. In a subsequent Phase II study utilizing IL-21 monotherapy for the management of metastatic melanoma, only 9 out of the 40 patients in the cohort (22.5%) showed partial objective responses, while 16 (40%) had stable disease (Petrella et al.,

2012). This discrepancy may be explained not only by inter-individual differences in immune status (e.g. availability of Signals 1 and 2 to stimulate T cell responses), but also the differential effects IL-21 has on the other immune cell types and subsets. For instance, IL-21 has been reported to induce immune-tolerance and inhibit Type 1-driven inflammation in vivo by inhibiting the maturation and promoting the apoptosis of DCs

(Brandt et al., 2003; Wan et al., 2013), which consequently impairs tumor antigen presentation to T cells. Moreover, by inducing the expression of RORγt and IL-23R, IL-

21 can also promote the differentiation, stabilization and expansion of Th17 cells (Zhou et al., 2007). While their role in tumorigenesis remains controversial, Th17 cells have been proposed to contribute to tumor development and metastasis through the production of pro-angiogenic IL-17 (Alizadeh et al., 2013; Singh et al., 2014). In view of the pleiotropic nature of IL-21, the anti-tumor effects of systemic IL-21 therapy may be counteracted by its pro-tumor effects.

The third γc-cytokine undergoing rigorous clinical testing is IL-15. In response to pro-inflammatory signals (e.g. TLR ligands, GM-CSF and Type I interferons) (Mattei et al., 2001; Neely et al., 2001), IL-15 expression is induced in multiple cell types, including activated innate immune cells (e.g. macrophages and DCs) (Carson et al., 1995; Mattei et al., 2001), as well as non-immune cells (e.g. stromal cells, epithelial cells and endothelial cells) (Cui et al., 2014). Secreted IL-15 binds with high affinity to its unique receptor, IL-

15Rα, which is expressed on activated APCs and stromal cells. Unlike the other γc- cytokines, secreted IL-15 has to be presented in a complex with IL-15Rα in order to gain biological activity. Secreted IL-15 may be presented in cis by directly binding the IL-

15Rα/IL-2R β/ γc trimer expressed on the same cell (Zanoni et al., 2013). However, T cells that do not express IL-15Rα instead rely on trans-presentation by APCs, a cell contact-dependent process involving ligation between the IL-15/IL-15Rα complex on

APCs and the IL-2R β/γc heterodimer on T cells (Burkett et al., 2004; Dubois et al.,

2002). Like IL-2, IL-15 stimulates the proliferation and expression of cytotoxic effectors in NK cells and T cells (Klebanoff et al., 2004; Zeng et al., 2005). However, IL-15 is regarded to be a more promising immune-stimulatory agent than IL-2, as IL-15 protects against activation-induced T cell death and does not potentiate Treg expansion (Berger et al., 2009; Brincks and Woodland, 2010; Saligrama et al., 2014). Due to the unique mechanism by which IL-15 is presented to effector NK cells and T cells, initial studies using systemic IL-15 administration was severely hampered by the challenges facing efficient IL-15 delivery. By itself, IL-15 protein is not only biologically inactive, but is also highly unstable with a short half-life (Bergamaschi et al., 2008). This is further exacerbated by the fact that anti-tumor effectors rely on the availability of neighboring

IL-15Rα-expressing APCs, which have to bind and trans-present IL-15 in a contact- dependent fashion. To overcome these difficulties, the field of immunotherapy has looked to IL-15/IL-15Rα fusion products to enhance the bioactivity and bioavailability of

IL-15 (Dubois et al., 2008; Stoklasek et al., 2006). Phase I/II clinical trials for one such

fusion product, ALT-803 (Altor Bioscience), has recently been launched in late 2014 for the treatment of metastatic melanoma and hematological cancers.

1.1.3 Antibody-based approaches

The profound brakes immune checkpoints put on anti-tumor immune responses have spurred the emergence of checkpoint blockade therapies targeting CTLA-4, PD-1 and LAG-3. CTLA-4 and PD-1 blockade are achieved by the systemic administration of monoclonal antibodies that inhibit cognate interaction with their respective ligands. As described in Chapter 1.1.1, CTLA-4 and PD-1 function at different phases of T cell activation. Therefore, anti-CTLA-4 and anti-PD-1 therapies exert different modes of action in restoring immune-competence and T cell activation: whereas anti-CTLA-4 mainly promotes priming of anti-tumor T cells in the tumor draining lymph nodes, anti-

PD-1 predominantly functions to restore the effector function of anti-tumor T cells within the tumor microenvironment. In contrast, the bifurcating effects of LAG-3 ligation on innate APCs and cytotoxic effector cells have precipitated the development of the LAG-

3-Ig fusion protein, which serves as a competitive inhibitor of endogenous cell-surface

LAG-3.

Two human anti-CTLA-4 monoclonal antibodies, ipilimumab (Bristol-Myers

Squibb) and tremelimumab (Pfizer), have been developed, and have received FDA approval for the therapy of a variety of solid tumors, including melanoma, hepatocellular carcinoma and lung cancer (Blank and Enk, 2014). Systemic anti-CTLA-4 treatment, when used alone or in combination with chemotherapy or immunotherapy, could produce durable responses and significantly prolong patient overall survival (Kirkwood et al., 15

2010; Ribas et al., 2013). A recent study reported that CTLA-4 checkpoint blockade is particularly effective for highly-mutated tumors that are enriched in TSAs, due to the ability of anti-CTLA-4 to augment endogenous anti-tumor effector T cells (Snyder et al.,

2014). However, multiple immune-related adverse events (irAEs) occurred in 80-95% of patients, with as many as a third of patients experiencing severe grade 3-4 drug-induced irAEs (O'Day et al., 2010; Weber et al., 2009). Moreover, some patients developed new tumor lesions or had increased tumor burdens after the initiation of anti-CTLA-4 therapy

(O'Day et al., 2010; Weber et al., 2009). These immune-related toxicities are possibly due to elevated, non-specific systemic inflammation occurring in peripheral lymphoid organs resulting from pan-inhibition of CTLA-4. The low safety profile of anti-CTLA-4 therapy therefore restricts its widespread clinical application. Limiting CTLA-4 blockade to selected patients who are most likely to benefit – such as those with TSA-enriched tumors (Snyder et al., 2014) – may enhance the therapeutic margin of this treatment.

In contrast to anti-CTLA-4 therapy, Phase I clinical trials have shown that antagonistic PD-1 and PD-L1 antibodies are generally well-tolerated and may restore immune-competence within the tumor microenvironment (Brahmer et al., 2012; Hamid et al., 2013; Robert et al., 2015), making PD-1 blockade a more feasible therapeutic option.

Although these therapies can stabilize disease progression and significantly prolong overall survival, patient responses are highly varied across cancer types: in the treatment of advanced solid tumors, objective response rates ranged from 6% for ovarian cancer to

50% for melanoma (Brahmer et al., 2010; Brahmer et al., 2012; Hamid et al., 2013;

Robert et al., 2015; Topalian et al., 2012). This is in part attributed to the fact that a large

proportion of human tumors do not express PD-L1 or PD-L2, rendering these antagonistic antibodies ineffective (Nomi et al., 2007; Topalian et al., 2012). However, even among PD-L1-expressing advanced melanomas – the malignancy that has thus far demonstrated the highest clinical responsiveness – only 40% of patients benefitted from anti-PD-1 therapy (Topalian et al., 2012). This strongly suggests the interplay of additional immune-barriers (e.g. TGF β) within the suppressive tumor microenvironment, and reiterates the need for alternative or complementary therapeutic strategies to optimally preserve CTL functional competence.

The function of LAG-Ig fusion protein, IMP321, is two-fold. First, IMP321 can engage MHCII on innate APCs, such as DCs, monocytes and macrophages, to promote their functional maturation. Second, IMP321 serves as a competitive inhibitor of endogenous LAG-3 expressed on activated T cells and NK cells, thus preserving the immune-competence of these anti-tumor effectors. Phase I and II clinical trials for

IMP321 – either as a single agent or in combination with chemotherapy – demonstrated that it is well-tolerated, eliciting little to no adverse events (Brignone et al., 2009;

Brignone et al., 2010). Monotherapy with IMP321 promoted the activation of circulating

CD8+ T cells and augmented long-lived CD8+ Tem generation in a cohort of patients with advanced renal cell carcinoma (Brignone et al., 2009). Although IMP321 alone could induce stable disease, none of the patients experienced objective responses

(Brignone et al., 2009), suggesting that it may be more efficacious as part of combinatorial therapies. Indeed, when combined with the first-line chemotherapeutic agent paclitaxel, IMP321 increased the objective response rate from 25% to 50% in

patients with metastatic breast cancer (Brignone et al., 2010). Clinical benefit in these patients was attributed to an enhanced proliferation and functional maturation of innate

APCs (e.g. monocytes and macrophages), in turn resulting in an increased expansion of

CD8+ T cells and NK cells (Brignone et al., 2010). In addition, preclinical studies have demonstrated that dual blockade of LAG-3 and PD-1 leads to more potent tumor regression than either agent alone (Jing et al., 2015; Woo et al., 2012), laying the foundation for a Phase I clinical trial investigating the co-inhibition of LAG-3 and PD-1 for metastatic solid tumors (NCT01968109).

1.1.4 Cell-based approaches

Cancer immunotherapy can also be achieved by directly targeting and boosting the function of specific immune cell types, particularly DCs, NK cells and T cells. These cell-based approaches often involve the expansion, activation and differentiation of immune cells under optimal conditions ex vivo, so that potent anti-tumor effectors can be generated and reinfused into the patient in great numbers.

As professional antigen-presenting cells, DCs represent a critical bridge between innate immunity and adaptive immunity: DCs are unique in their abilities not only to efficiently cross-present exogenously-derived tumor antigens to anti-tumor T cells, but also to orchestrate T cell responses by delivering either inflammatory or tolerogenic signals. These features have motivated the development of two main strategies of DC- targeted vaccines for cancer immunotherapy. Ultimately, the goal of both DC-targeted vaccination strategies is to elicit strong and durable anti-tumor CD8+ T cell responses in the patient. In the first approach, ex vivo-activated and matured DCs are loaded with 18

tumor antigens and then reinfused intravenously into the patient. Clinical trials have successfully demonstrated the safety and efficacy of this approach in prolonging the survival of prostate cancer patients, leading to FDA approval of Sipuleucel-T for the treatment of metastatic prostate cancer (Higano et al., 2009; Kantoff et al., 2010).

Sipuleucel-T ‒ comprised of autologous DCs treated ex vivo with GM-CSF and pulsed with the human prostatic acid phosphatase (PAP) TAA ‒ has recently been reported to elicit clinical benefit by enhancing the recruitment and proliferation of anti-tumor effector T cells within the prostate tumor microenvironment (Fong et al., 2014). In the second method, patients are injected with tumor antigens conjugated to monoclonal antibodies against DC-specific endocytic receptors (e.g. Fcγ receptors or scavenger receptors), resulting in the selective targeting and uptake of tumor antigens by DCs in vivo (Caminschi et al., 2009; Hawiger et al., 2001). Nevertheless, both approaches have their drawbacks. The first challenge is tumor antigen selection. To prevent autoimmunity,

DC-targeted vaccines should ideally deliver tumor-specific antigens (TSAs); however,

TSAs are often specific and unique to each patient and would therefore require a tailored process of antigen identification and isolation for each patient (Linnemann et al., 2015;

Tran et al., 2014; Yadav et al., 2014). On the other hand, while non-mutated tumor- associated antigens (TAAs) are shared among most patients, targeting TAAs runs a high risk of autoimmunity due to on-target, off-tumor damage. Second, ex vivo-generated DC vaccines are laborious and expensive to produce: as a form of personalized medicine,

DCs will have to be isolated and prepared from each individual patient for autologous reinfusion. However, current procedures used for the in vitro isolation and expansion of

DCs are time-consuming and costly, thus undermining the feasibility of this approach.

Third, targeting conjugated antigens to DCs in vivo requires co-administration of a suitable TLR ligand adjuvant, such as poly-IC (Kastenmuller et al., 2011; Wille-Reece et al., 2006), which delivers a “danger” signal to endogenous DCs. Adjuvants are crucial for anti-tumor efficacy, because in the absence of a “danger” signal, DCs presenting the targeted antigen remain immature and become tolerogenic, instead of acquiring anti- tumor functions (Hawiger et al., 2001). On the flipside, adjuvants can elicit systemic, non-specific immune-activation: in non-human primates, intradermal injection of adjuvants alone resulted in elevated numbers of circulating neutrophils and monocytes, as well as increased pro-inflammatory cytokine concentrations in the serum (Kwissa et al.,

2012). The low safety profile and high risk of auto-inflammation are therefore stumbling blocks in the clinical development of DC-targeted cancer vaccines.

The adoptive transfer of ex vivo-expanded NK cells has also been explored for cancer immunotherapy. NK cells are attractive candidates as they can secrete anti-tumor cytokines (e.g. IFNγ and TNFα) (Fauriat et al., 2010), and are equipped to directly lyse tumor cells through death receptor ligation (e.g. FasL and TRAIL), granzyme release and antibody-dependent cytotoxicity (ADCC) (Alderson and Sondel, 2011; Smyth et al.,

2005). However, the clinical utility of NK cell for ACT has been limited by (i) the nature of NK cell target recognition, (ii) poor in vivo expansion and survival and (iii) susceptibility to tumor-induced evasion and functional suppression. The “missing self” model aptly describes how NK cells recognize tumor cells. NK cells express inhibitory killer-immunoglobulin-like receptors (KIRs) that when bound to self-MHCI molecules,

prevent NK cell-mediated cytolysis (Vilches and Parham, 2002); however, cells that lose surface MHCI expression can no longer engage these inhibitory KIRs, resulting in NK cell activation and target cell killing. As target recognition by NK cells requires either the absence of self-MHCI or mismatched MHCI on tumor cells, ACT with autologous NK cells is not efficacious (Knorr et al., 2014). To enhance the efficiency of tumor cell recognition, allogenic NK cells from MHC-haploidentical donors may instead be transferred into patients. However, the in vivo persistence of these transferred allogenic

NK cells still remains poor. To support their in vivo expansion and survival, treatment of acute myeloid leukemia patients with allogenic NK cells had to be supplemented by co- therapy with high-dose IL-2 and extensive lymphodepleting regimens in order to achieve clinical efficacy (Miller et al., 2005). Such an approach, too, has its drawbacks, as high- dose IL-2 administration promotes the expansion of Tregs (Sim et al., 2014), which can impair NK cell cytokine secretion and cytotoxicity (Pedroza-Pacheco et al., 2013). To overcome this, IL-2 co-therapy may either be used in conjunction with Treg-depleting regimens or be substituted with IL-15. Nevertheless, even when multiple doses of allogenic NK cells with IL-15 were administered into a cohort of 16 non-small cell lung cancer patients, only 2 had partial responses and 6 had stable disease (Iliopoulou et al.,

2010). This low clinical efficacy is mainly attributed to the evasion of NK cell-mediated killing (Doubrovina et al., 2003; Wu et al., 2004) and functional suppression of NK cells by the tumor microenvironment (Vitale et al., 2014), with the latter posing an exceptional barrier particularly in the treatment of solid tumors. Although NK cells are attractive vehicles for ACT, these caveats have largely restricted their widespread clinical use.

The third immune cell type utilized for ACT is T cells. T cells are ideal candidates for ACT, due to their (i) ability to recognize specific tumor antigens, (ii) cytotoxic effector functions and (iii) potential to generate immunological memory for long-lasting anti-tumor immune-surveillance. Conventionally, ACT using autologous T cells relies on the isolation of tumor-infiltrating T cells (TILs) from the excised tumor, massive ex vivo expansion, followed by the screening and selection of T cells bearing reactivity towards autologous tumor cells(Rosenberg et al., 2008). Clinical trials demonstrated that this approach, when combined with high-dose IL-2 and lymphodepleting chemotherapy, could induce tumor regression in over 50% of patients with refractory metastatic melanoma (Dudley et al., 2002; Rosenberg and Dudley, 2004). However, melanoma patients who responded to therapy also exhibited signs of autoimmune melanocyte destruction in normal skin (i.e. vitiligo) and the eye (i.e. uveitis) (Dudley et al., 2002), highlighting lack of precision and tumor-specificity in T cell selection. In addition to these on-tumor off-target side effects, it remains extremely laborious and costly to generate sufficient numbers of anti-tumor autologous T cells for reinfusion, even with current rapid T cell expansion procedures: the entire process of isolating, selecting and expanding tumor-reactive T cells from TILs ex vivo requires a total of 4-5 weeks (Tran et al., 2014), in order to obtain a minimum of 10 10 T cells for therapeutic efficacy (Tran et al., 2014). Furthermore, functional anti-tumor TILs may not always be recoverable from all patients for ex vivo culture. This is because chronic antigen stimulation or the immune-suppressive tumor microenvironment may push tumor-specific TILs into a permanent state of anergy or hyporesponsiveness. Finally, not all patients responded to

autologous T cell ACT, as these reinfused T cells remain vulnerable to immune- suppression by the tumor microenvironment (Fourcade et al., 2010; Motz et al., 2014;

Thomas and Massague, 2005). Nevertheless, autologous T cells have proven to show the greatest clinical efficacy and are thus far the most popular candidates for ACT.

1.2 CD8+ T cells in cancer immunotherapy

CD8+ T cells are a unique population of immune cells equipped with antigen specificity, cytotoxic effector machinery and potential for memory differentiation. They are therefore well-poised not only to seek out and eradicate malignant cells, but also to mediate long-lasting immune-surveillance against tumor relapse. To fully harness and manipulate these traits for optimal ACT cancer immunotherapy, we need to first understand the molecular mechanisms governing the fate and function of CD8+ T cells.

1.2.1 Immune-protection conferred by effector and memory CD8+ T cells

Upon cognate antigen encounter in lymph nodes, naïve CD8+ T cells become activated, undergo clonal expansion and begin a process of functional differentiation.

During this process, CD8+ T cells become armed with cytotoxic effector functions and acquire the capacity to migrate to peripheral sites where they eradicate target cells, such as aberrant or virus-infected cells. Following clearance of their targets, effector CD8+ T cells undergo a dramatic contraction phase, in which the vast majority of terminally- differentiated short-lived effector cells (SLECs) die by apoptosis, leaving behind a mere

5-10% memory precursor effector cells (MPECs) to form a long-lived memory pool

(Tm) (Araki et al., 2009; Kaech et al., 2002; Miller et al., 2008).

Effector and memory CD8+ T cells are qualitatively and phenotypically distinct.

KLRG1 hi IL-7Rαlow SLECs exhibit superior cytotoxic potency and robustly express multiple cytotoxic mediators, such as IFNγ and granzyme B proteins, which can be immediately released for rapid target cell lysis (Corazza et al., 2000; Kaech et al., 2002;

Peixoto et al., 2007); however, due to their short-lived nature and limited proliferative capacity, protective immunity mediated by SLECs quickly wanes. In contrast,

KLRG1 low IL-7Rαhi CD8+ Tm cells are not immediately cytolytic, as they do not constitutively express cytolytic proteins and therefore require some time following antigenic restimulation to re-acquire effector functions (Kaech et al., 2002). In spite being functionally latent, it has been shown that CD8+ Tm cells confer greater protection than

SLECs against rechallenge: the former’s slight delay in effector acquisition is outweighed by their enhanced survival and proliferative capacity, which provide a numerical advantage for the effective clearance of recurring infections (Kaech et al., 2002; Peixoto et al., 2007).

CD8+ Tm cells carry out long-term immune-surveillance, thus safeguarding the host against pathogenic reinfection or tumor recurrence. Based on their self-renewing capacity, homing properties and anatomical localization, CD8+ Tm cells can be further sub-categorized into CD44 + CD62L + CCR7 + CD127 + central memory (Tcm), CD44 +

CD62L - CCR7 - CD127 + effector memory (Tem) and the more recently-characterized

CD44 + CD69 + CD103 + resident memory (Trm) cells (Mueller et al., 2013). The functional properties of Tcms and Tems, as well as their complementary roles in host immune protection have been well-characterized. Tcms recirculate between secondary

lymphoid organs, scanning for antigens that drain to the lymph nodes and spleen

(Sallusto et al., 1999). On the other hand, Tems circulate between the blood and peripheral non-lymphoid tissues (e.g. the skin and mucosal sites), where they patrol for antigens at potential sites of pathogen reinfection (Masopust et al., 2001; Sallusto et al.,

1999). Whereas Tcms have an enhanced potential to self-renew and undergo extensive secondary proliferation, Tems have limited expansion capacity (Wherry et al., 2003).

Tems, however, can respond more rapidly to antigenic rechallenge, as they are capable of immediate cytotoxicity (Olson et al., 2013); on the other hand, Tcms exhibit delayed recall cytotoxic responses, but retain the potential to self-renew and differentiate into

Tems (Sallusto et al., 1999).

The third CD8+ Tm subset, Trms, are comparatively less well-characterized.

Unlike Tcms and Tems that are in constant recirculation, Trms reside permanently in peripheral sites even after antigen clearance (Jiang et al., 2012; Schenkel et al., 2013). As

Trms are anatomically located at common entry sites of pathogen invasion, they are functionally poised as the frontline sentinels that detect and respond to recurring pathogens. In fact, Trms are the most effective CD8+ Tm subset in controlling viral reinfection at peripheral sites (Jiang et al., 2012). The role of Trms in executing immune- surveillance is multi-faceted. Trms can directly kill their targets by up-regulating cytotoxic mediators, thereby restricting the local pathogen load and limiting the spread of infection (Jiang et al., 2012). In the meantime, this provides a temporal window for circulating Tm cells (i.e. Tcms and Tems) to encounter the pathogen, undergo reactivation and be recruited to the site of infection. In addition, Trms can indirectly

promote pathogen clearance by initiating a “tissue-wide state of pathogen alert”, which amplifies the danger signal and orchestrates both innate and adaptive immune responses

(Ariotti et al., 2014; Schenkel et al., 2014; Schenkel et al., 2013). Trms achieve this by rapidly secreting a myriad of cytokines and chemokines that recruit circulating Tm cells and B cells, induce the functional maturation of local DCs and activate the killing machinery of local cytotoxic cells (e.g. NK cells and Trms) (Ariotti et al., 2014; Schenkel et al., 2014). These findings suggest that the presence of anti-tumor CD8+ Trms may be critical for enforcing long-lasting immune-surveillance.

That ACT necessitates the reinfusion of a tremendous number of tumor-reactive T cells (~10 10 ) for clinical efficacy underscores the ongoing numbers game between tumor cells and anti-tumor effectors. Especially in the case of solid tumors, reinfused anti-tumor

T cells face the overwhelming challenges of homing into extravascular tumor masses, and overcoming local immune-suppression within the tumor microenvironment. To build this arsenal of potent anti-tumor effectors, the reinfused CD8+ T cells should ideally be a multi-functional, heterogeneous population comprised of robustly-cytotoxic tissue- homing effectors that can immediately be deployed to eliminate malignant cells, as well as disseminated multi-potent memory cells that can survey for tumors and regenerate effectors to protect against tumor relapse.

1.2.2 Anti-tumor effector functions of CD8+ T cells

Upon antigen recognition, activated CD8+ T cells can elicit target cell lysis through direct and indirect killing mechanisms: (i) Fas:FasL death receptor-mediated killing, (ii) cytotoxic granule delivery and (iii) cytokine secretion. Synergism between 26

these multiple complementary cytolytic pathways is required to execute potent tumor cell killing, and to minimize the selection and outgrowth of heterogeneous, immune-evasive variants.

While Fas is constitutively expressed by most cell types, FasL expression is predominantly restricted to activated T cells and NK cells (Leithauser et al., 1993).

Following priming in the lymph nodes, protein stores of FasL are synthesized and stored in intracellular granules of activated CD8+ T cells (Kassahn et al., 2009). When these

CD8+ T cells encounter target cells expressing their specific antigen, antigen-induced

TCR activation signals intracellular FasL to be rapidly transported to the cell membrane

(Kassahn et al., 2009). Interaction between FasL on effector CD8+ T cells and Fas on target cells activates the death receptor-induced caspase pathway, resulting in target cell apoptosis (Nagata and Golstein, 1995). Tumor-specific effector CD8+ T cells deficient in

FasL are defective in controlling tumor growth, demonstrating the importance of

Fas:FasL interactions in tumor clearance (Listopad et al., 2013). Nonetheless, human cancers frequently evade FasL-induced apoptosis by either altering Fas surface expression or expressing a loss-of-function Fas protein (Lee et al., 1999; Viard-Leveugle et al., 2003): Fas expression has been reported to be down-regulated in as many as 90%, and completely lost in 24% of human lung cancers (Lee et al., 1999; Viard-Leveugle et al., 2003).

Effector CD8+ T cells may alternatively destroy their targets through the delivery of granzyme-containing cytotoxic granules. Granzymes are a family of pro-apoptotic serine proteases that are expressed almost exclusively by activated CD8+ T cells and NK

cells. In activated CD8+ T cells, granzymes are stored intracellularly in acidified cytotoxic granules, together with monomers of the pore-forming protein perforin. Such compartmentalization not only physically separates granzymes from their cytosolic substrates within the CD8+ T cell, but the low pH also keeps granzymes and perforin monomers functionally inactive (Lopez et al., 2012). During target cell recognition, an immunological synapse formed at the region of TCR-pMHC contact serves as a focal point for the unidirectional secretion of cytotoxic granules. Cytoskeletal remodeling in activated T cells polarizes the secretory machinery, allowing the directional release of cytotoxic granules towards the immunological synapse (Beal et al., 2009). This unidirectional degranulation is critical to ensure lysis of the target cell, while sparing normal bystander host cells. Upon granule release and exposure to neutral pH, perforin oligomerizes to form a pore in the membrane of the target cell, facilitating the delivery of functionally-active granzymes into the cytosol (Baran et al., 2009). Among the granzymes, granzyme B is the most well-characterized and most highly-expressed in effector CD8+ T cells. Granzyme B activates apoptosis by proteolytically cleaving multiple crucial intracellular pro-apoptotic substrates (e.g. the executioner caspase-3 and

Bid) to their active forms, which in turn trigger apoptosis and cell death (Barry et al.,

2000; Quan et al., 1996). Cytotoxic granule delivery by effector CD8+ T cells represents a key safeguard in anti-tumor immune-surveillance. Compared to wildtype mice, perforin-deficient mice have an increased propensity for developing spontaneous lymphoma, primarily due to defective clearance by CD8+ T cells (Smyth et al., 2000).

While a clear role has been established for perforin, a point of contention arises as to the

necessity of granzymes in tumor control: when adoptively transferred into melanoma- bearing mice, CD8+ T cells doubly-deficient in granzymes A and B were as competent as their wildtype counterparts in tumor rejection (Davis et al., 2001), suggesting that granzymes A and B are not required for tumor clearance. Paradoxically, tumors that overexpress serpins ‒ serine protease inhibitors that inhibit granzyme proteolysis ‒ are resistant to CD8+ T cell-mediated killing both in vitro and in vivo (Medema et al., 2001).

Moreover, human lymphomas commonly overexpress the serpin PI9 as an immune- evasive mechanism to escape CD8+ T cell destruction (Bladergroen et al., 2002; Bossard et al., 2007), and PI9 expression correlates positively with disease malignancy

(Bladergroen et al., 2002). This discrepancy may due to functional redundancy between various cytotoxic effector pathways that are at play in the clearance of immunogenic tumors (Waterhouse et al., 2006); however, when faced with immune-evasive or immune-suppressive tumors that can resist certain arms of cytotoxic killing (e.g. FasL resistance), granzyme-induced apoptosis comes to the forefront of anti-tumor defense.

In addition to direct targeted destruction by FasL or cytotoxic degranulation, effector CD8+ T cells can also indirectly promote target cell lysis. This is achieved by secretion of the immune-activating cytokine IFNγ, which coordinates diverse anti-tumor effects. Firstly, IFNγ changes the proteasome structure in APCs, such that proteasomal cleavage products become skewed towards peptide epitopes that can be loaded onto

MHCI. Under homeostatic conditions, APCs express the constitutive proteasome, which contains a 20S proteolytic core comprising of the β1 (δ), β2 (MB1) and β5 (Z) subunits

(Kloetzel, 2001). The constitutive proteasome subunits preferentially cleave poly-

ubiquitinated proteins into peptides with hydrophilic termini, which are disfavored from binding the hydrophobic pockets at the ends of the MHCI groove (Toes et al., 2001).

However, IFNγ induces the exchange of the β1 (δ), β2 (MB1) and β5 (Z) subunits to the

β1i (LMP2), β2i (MECL-1) and β5i (LMP7) subunits, respectively, forming the immunoproteasome (Kloetzel, 2001). The immunoproteasome has a substrate binding pocket with a distinct structure and activity, which favors the generation of peptides with hydrophobic termini for preferential docking into MHCI (Huber et al., 2012). In fact, the immunoproteasome increases not only the quantity (Schwarz et al., 2000), but also the diversity of peptides that are presented by MHCI (de Verteuil et al., 2010), enhancing the breadth of antigen recognition and activation of CD8+ T cells. In this way, IFNγ profoundly impacts the epitope repertoire available for MHCI antigen presentation.

Secondly, IFNγ can up-regulate MHCI expression on both APCs and tumor cells.

Increased MHCI expression on APCs augments tumor antigen presentation, and consequently promotes the priming of anti-tumor CD8+ T cells (Brucet et al., 2004).

More importantly, IFNγ-induced MHCI expression on tumor cells facilitates immune- recognition by anti-tumor CD8+ T cells, rendering tumor cells more susceptible to immune clearance (Seliger et al., 1997; Tajima et al., 2004); this aspect is especially crucial in tumors that evade CD8+ T cell recognition by down-regulating surface MHCI

(Tajima et al., 2004). Thirdly, IFNγ generates a Type 1 immune milieu within the tumor microenvironment for optimal tumor elimination. IFNγ augments macrophage activation, resulting in enhanced direct tumor cell killing by macrophages (e.g. by increasing the expression of FasL and secretion of reactive oxygen species), and elevated production of

IL-12 that further reinforces Th1 differentiation (Hu et al., 2008). IFNγ concurrently promotes Type 1 differentiation of CD8+ and CD4+ T cells by inducing expression of the transcription factor T-bet. T-bet reinforces Type 1 polarization of T cells by driving IFNγ and IL-12Rβ expression, and is capable of re-directing Th2 cells towards the Th1 lineage

(Szabo et al., 2000). Lastly, IFNγ-Stat1 signaling directly inhibits tumor growth by inducing the expression of cell cycle inhibitors (e.g. Cdk2 and Cdk4) (Bromberg et al.,

1996; Chen et al., 2000; Mandal et al., 1998), as well as apoptotic mediators (e.g. Fas and caspase-1) (Xu et al., 1998) in tumor cells. These anti-tumor effects of IFNγ speak to its importance in tumor rejection and surveillance: in mouse tumor models of methylcolanthrene -induced sarcoma, IFNγR-deficient or IFNγ-deficient immune- competent mice both showed increased frequencies of tumor occurrence accompanied by accelerated tumor progression (Kaplan et al., 1998; Street et al., 2001). These studies highlighted that the other arms of an intact immune compartment cannot compensate for the lack of IFNγ, underscoring the non-redundancy and indispensability of IFNγ in tumor rejection.

1.2.3 Transcriptional regulation of CD8+ T cell effector functions and memory generation

The function and fate of activated CD8+ T cells are determined by an intricate interplay and balance between upstream master transcription factors, including T-bet,

Blimp-1 and Eomes.

The T-box transcription factors, T-bet and Eomes, are two key players that regulate both the effector functions and memory differentiation of activated CD8+ T

cells. While absent in naïve CD8+ T cells, expression of these T-box factors are rapidly induced upon T cell activation and maintained in both effector and memory CD8+ T cells

(Intlekofer et al., 2005). In effector CD8+ T cells, T-bet and Eomes are compensatory and essential transcriptional factors enforcing a Type 1 program that instructs their differentiation into highly-potent killer cells – T-bet and Eomes drive the expression of

Type 1 cytotoxic effector molecules (e.g. granzyme B, perforin and IFNγ) for the eradication of malignant cells, while simultaneously repressing the acquisition of an unproductive Type 17 program that targets extracellular pathogens (Cruz-Guilloty F,

Interestingly, although T-bet and Eomes synergize in effector CD8+ T cells to impart full-fledged cytotoxic functions, the interplay between these two transcription factors is more complex in memory CD8+ T cells. It is believed that immediately following CD8+ T cell activation, the effector-versus-memory transcriptional program is instructed by the counterbalance between T-bet and Eomes expression. During CD8+ T cell priming, a highly-inflammatory, IL-12-rich cytokine milieu robustly induces T-bet, resulting in the generation of IL-7Rα low effector CD8+ T cells that preferentially undergo terminal effector differentiation to SLECs; reciprocally, effector CD8+ T cells expressing low levels of T-bet could up-regulate IL-7Rα to receive IL-7-transduced pro-survival signals, and preferentially differentiated to MPECs with memory potential (Joshi et al.,

2007). Although T-bet represses IL-7Rα expression (Intlekofer et al., 2007), it induces

IL-2Rβ expression to confer effector CD8+ T cells with IL-15 responsiveness(Li et al.,

2011); therefore, moderate levels of T-bet are still required to support CD8+ T cell effector-to-memory transition (Intlekofer et al., 2005; Juedes et al., 2004). High levels of

Eomes, on the other hand, are driven by IL-2 and promote effector CD8+ T cell differentiation to an MPEC phenotype for enhanced memory CD8+ T cell generation

(Banerjee et al., 2010; Hinrichs et al., 2008; Rao et al., 2010). Immediately following

CD8+ T cell priming, T-bet and Eomes thus exert opposing effects in SLEC-versus-

MPEC differentiation. After the establishment of a memory pool, both T-bet and Eomes support long-term survival of memory CD8+ T cells by driving the optimal expression of

IL-2Rβ, allowing the transduction of pro-survival signals through IL-15 (Intlekofer et al.,

2005). However, they have antagonistic roles in Tem-versus-Tcm fate determination: in memory CD8+ T cells, high T-bet expression promotes terminal effector differentiation to Tems at the expense of Tcms (Intlekofer et al., 2007), whereas high levels of Eomes favors development to Tcms (Banerjee et al., 2010). A balance between T-bet and Eomes is therefore required to establish a heterogeneous population of CD8+ T cells, in which

Tems provide a rapid recall cytotoxic response, while Tcms serve as a highly- proliferative pool for self-renewal.

Operating further upstream in this transcriptional hierarchy is the master regulator

Blimp-1. Unlike the T-bet and Eomes transcription factors that directly drive gene expression, Blimp-1 serves as an epigenetic remodeling factor: Blimp-1 binds to specific

DNA sequences and recruits chromatin-modifying enzymes to alter the accessibility of targeted gene regulatory regions (Lu et al., 2014; Yu et al., 2000). Although originally

identified as a master regulator of plasma B cell differentiation (Turner et al., 1994),

Blimp-1 has subsequently been found to play divergent roles in controlling the function and fate of activated CD8+ T cells. Following antigen stimulation, the induction of

Blimp-1 expression in effector and effector-memory CD8+ T cells is essential for turning on the transcriptional program of cytotoxic effectors (Kallies et al., 2009; Rutishauser

RL, 2009; Shin H, 2009). In mouse models of acute and chronic virus infection, Blimp-1- deficient primary effector and Tem CD8+ T cells resulted in delayed viral clearance

(Kallies et al., 2009; Rutishauser RL, 2009), highlighting the importance of Blimp-1 in promoting cell-mediated immunity during both primary and recall responses. The enhancement of cytotoxic effector responses by Blimp-1 is three-fold. First, by regulating the expression of chemokine receptors (e.g. CCR7 and CCR5), Blimp-1 enables activated

CD8+ T cells to migrate from the draining lymph nodes to peripheral tissues where infected or malignant cells are located (Kallies et al., 2009). Second, Blimp-1 supports the cytotoxic transcriptional profile in CD8+ T cells, by enhancing the expression of multiple cardinal effector molecules (e.g. granzymes and perforin), as well as the transcription factor T-bet (Kallies et al., 2009; Rutishauser et al., 2009). Third, during the early phases of CD8+ T cell activation, Blimp-1 prevents the acute exhaustion of CD8+ effector T cells by silencing the transcription of co-inhibitory PD-1 (Lu et al., 2014).

Following CD8+ T cell activation, PD-1 up-regulation precedes the induction of Blimp-1.

While PD-1 serves as a negative feedback mechanism that prevents hyper-inflammation, the quenching of effector CD8+ T cell functions so early on in the immune response hampers effective target clearance. In this regard, Blimp-1 is essential for down-

regulating PD-1 expression in effector CD8+ T cells to preserve their cytotoxicity and enforce protective immunity (Lu et al., 2014). Therefore, Blimp-1 is a crucial driver of cytotoxic functions in effector CD8+ T cell.

During the late effector and memory phases of the CD8+ T cell response, however, the impact of Blimp-1 is more controversial. High Blimp-1 expression in late effector CD8+ T cells has been reported to undermine memory differentiation for two reasons. First, persistently elevated levels of Blimp-1 turn on a transcriptional program that favors terminal differentiation of effector CD8+ T cells to KLRG1 high IL-7R low

SLECs, at the expense of MPECs and long-lived Tcm cells (Kallies et al., 2009;

Rutishauser et al., 2009; Shin et al., 2013). This is because Blimp-1 (i) deprives late effector CD8+ T cells of pro-survival signals by directly repressing the gene expression of IL-2, IL-2Rα, the co-stimulatory molecule CD27 (Martins et al., 2008; Shin et al.,

2013); and (ii) drives late effector CD8+ T cells into senescence by impairing their DNA replication and repair machinery (Ji et al., 2011). Blimp-1-expressing cells that transition into the memory phase therefore preferentially adopt the Tem lineage that is characterized by efficient cytotoxic recall responses, but low proliferative capacity

(Rutishauser et al., 2009). Second, elevated levels of Blimp-1 has also been reported to cause CD8+ T cell exhaustion, due to the concomitant induction of multiple inhibitory molecules, such as PD-1, LAG-3 and 2B4 (Shin et al., 2009). As a result, Blimp-1- deficient CD8+ T cells preferentially differentiate into MPECs that go on to form a larger

Tcm-enriched memory pool bearing reduced expression of inhibitory surface markers

(Shin et al., 2009). Nevertheless, compared to wildtype CD8+ T cells, Blimp-1-deficient

CD8+ T cells are ineffective at controlling chronic viral infections, as their enhanced proliferative capacity could not compensate for their diminished granzyme B expression and cytotoxicity (Rutishauser et al., 2009; Shin et al., 2009). Intriguingly, Blimp-1- haploinsufficient CD8+ T cells showed little signs of exhaustion while retaining substantial cytotoxic potency, making them superior to their wildtype counterparts in clearing chronic viral infections (Shin et al., 2009). This not only demonstrates that

Blimp-1 is critical for effective CD8+ T cell killer functions, but also suggests that intermediate levels of Blimp-1 ‒ low enough to prevent exhaustion, but high enough to drive cytotoxic effector functions ‒ can optimally confer immune protection.

1.3 Chimeric antigen receptors (CARs): Novel gene therapy tools for ACT

Chimeric antigen receptors (CARs) are novel gene therapy tools that can be used to redirect host immunity against specific tumor antigens. Comprised of a single chain antibody fragment (scFv) fused to intracellular T cell signaling domains, CARs can couple high-affinity antigen recognition to efficient T cell activation (Eshhar et al., 1993).

To date, CARs bearing scFv domains specific for a variety of hematological and solid tumor antigens have been developed. These target antigens include tissue-specific antigens such as CD19 (for leukemia and lymphoma), TAAs such as Her2 (for breast and colorectal cancers) and Muc1 (for breast, ovarian and prostate cancers), as well as TSAs such as EGFRvIII (for glioblastoma) (Essand and Loskog, 2013; Kershaw et al., 2014;

Wilkie et al., 2008). Intracellular signaling moieties of CARs invariably combine CD3ζ

with one or more co-stimulatory domains (e.g. 4-1BB, CD28 and OX40) (Lipowska-

Bhalla et al., 2012).

1.3.1 Utility of CAR T cell-based ACT

Advances in the field of T cell engineering have enabled the integration of CARs with ACT: autologous lymphocytes isolated from peripheral blood can now be transduced ex vivo with CAR-encoding viral vectors, and then reinfused into the patient

(Kalos et al., 2011; Scholler et al., 2012). In this way, polyclonal peripheral T cells are genetically modified with specificity for the targeted antigen, and their effector responses can be redirected towards antigen-expressing tumor cells. In fact, CARs have revolutionized the utility of T cell-based ACT cancer immunotherapy in several ways.

First, as CARs incorporate antigen recognition and co-stimulatory domains in a single entity, both Signals 1 and Signal 2 can be delivered simultaneously to promote

CAR T cell activation and persistence (Maher et al., 2002). This offers an advantage over conventional ACT based on unmodified TILs, where TCR activation (Signal 1) is uncoupled from co-stimulation, and thus necessitates both antigen recognition and cognate co-stimulatory ligands for optimal effector function and survival (Curran et al.,

2011; Hernandez-Chacon et al., 2011; van Elsas et al., 1999). CARs therefore allow T cell functional competence to be preserved in vivo, especially in the tumor-bearing state where antigen presentation is impaired or co-stimulatory ligands are limiting.

Second, the process of ACT has been made more streamlined and efficient, since

CAR-modified T cells can now be generated from a readily-accessible and abundant source of peripheral blood lymphocytes (PBLs). This greatly mitigates the difficulties in 37

the isolation and expansion of TILs from surgically excised tumor fragments, considering that TIL-derived T cell products require an ex vivo culture period of 4-5 weeks (Tran et al., 2014), whereas therapeutic doses of CAR T cells derived from autologous PBLs can be manufactured within 10-12 days (Kalos et al., 2011).

Third, antigen recognition by CARs is not MHC-restricted. As CD8+ CAR T cells are invulnerable to MHCI-downregulation, tumor cells exploiting this common pathway of immune-evasion can still be efficiently recognized and cleared. Moreover, unlike TCR gene transfer approaches that necessitate MHC compatibility, the MHC- independence of CARs allows them to be administered across all patient MHC haplotypes.

The promise of CAR therapy is evident from the success of the anti-CD19-CAR targeting B cell malignancies. In patients with advanced or refractory B cell malignancies, therapy with autologous anti-CD19-CAR T cells could attain durable remissions in as many as 80%-100% of patients for up to 15 months ongoing, especially when aided by preconditioning chemotherapy and IL-2 cytokine support (Brentjens et al.,

2013; Kochenderfer et al., 2012). Immediately following patient reinfusion, anti-CD19-

CAR T cells expanded extensively in vivo and induced pro-inflammatory cytokine and chemokine secretion (Kalos et al., 2011), demonstrating potent short-term anti-tumor effector responses. Importantly, up to 6 months post-transfer, anti-CD19-CAR T cells remained detectable in the peripheral blood and bone marrow, expressed cell surface markers characteristic of memory T cells and were capable of CD19-specific cytotoxic recall responses (Kalos et al., 2011), indicating that CAR therapy can generate a

functional memory CAR T cell pool to mediate long-term protection. Taken together, clinical trials with the anti-CD19-CAR are a testament to the immense potential of CAR

T cell-based ACT for cancer immunotherapy.

1.3.2 Current challenges facing CAR T cell-based ACT

In spite of their promise, the excitement surrounding CAR T cell-based ACT has been tempered by several challenges that limit their widespread clinical use and therapeutic efficacy.

First, the exceptional affinity and unprecedented sensitivity of CARs for their antigen – mostly tumor-associated and not tumor-specific – poses a risk of bystander damage to normal tissues. In recent clinical trials of CAR therapy, cytokine storms and massive tissue damage have led to serious adverse events, and even patient deaths

(Brentjens et al., 2010; Kalos et al., 2011; Lamers et al., 2006; Morgan et al., 2010). This calls for further carefully-designed dose-escalation studies to minimize the quantities of

CAR T cells used, while still maintaining clinical efficacy.

Second, despite its unprecedented efficacy in hematological malignancies, CAR therapy has met with limited success in the treatment of solid tumors. Even for glioblastoma, in which the tumor-specific EGFRvIII-CAR is utilized, CAR therapy delayed tumor progression but was not completely curative (Choi et al., 2014). This partial anti-tumor efficacy in vivo may be attributed to the suppression of CAR T cell effector responses by a myriad of immune barriers within the tumor microenvironment, such as TGFβ and co-inhibitory receptors (e.g. PD-1 and LAG-3) (Ahmadzadeh M, 2005;

Moon et al., 2014; Thomas and Massague, 2005; Zhang et al., 2013a). This highlights 39

that the maximal provision of Signals 1 and 2 still fails to sufficiently preserve CAR T cell immune-competence within profoundly suppressive tumor microenvironments, reiterating the need to engineer functionally resilient CAR T cells that are capable of overcoming immune barriers in vivo.

Third, how reinfused CAR T cells cross-talk with the endogenous host immune system, as well as the repercussions on long-term anti-tumor protection, remain poorly characterized. The exquisite specificity of CARs for a single targeted antigen initially raised concerns: restricting CAR T cell reactivity to a single antigen within a heterogeneous tumor may potentially lead to cancer immune-editing, breakdown of CAR

T cell-mediated immune-surveillance, and eventual immune escape of antigen-loss variants. Several recent reports, however, have cast doubts on these postulations: in mouse models of lymphoma, ovarian cancer and glioblastoma, targeted therapy with tumor-specific CAR T cells not only mediated the clearance of primary tumors expressing the targeted antigen, but more importantly, rendered mice resistant to subsequent challenge with tumors lacking the targeted antigen (Barber et al., 2008;

Sampson et al., 2014; Spear et al., 2013a; Zhang et al., 2012). These findings support the notion that albeit being uniquely specific for a single antigen, CAR T cell therapy can protect against heterogeneous tumors. The underlying mechanisms and endogenous immune cell populations mediating this broad-spectrum, long-lasting protection, however, have not been defined. Elucidating the long-term impact of CAR T cell therapy will therefore help determine its therapeutic durability and guide its clinical application for the treatment of heterogeneous tumors.

1.4 MicroRNAs as targets for T cell immune-modulation

MicroRNAs (miRNAs), a group of evolutionarily-conserved single-stranded non- coding RNAs (20-24 nucleotides in length), are key regulators of gene expression in plants and animals (Carrington and Ambros, 2003). MiRNA genes encoded in the genome are transcribed into primary miRNA (pri-miRNA) transcripts, which undergo sequential cleavage and processing by the Drosha microprocessor complex and the Dicer

RNAse to generate mature miRNAs for loading onto the RNA-induced silencing complex (RISC). By guiding RISC to complementary mRNA sequences, mature miRNAs direct the silencing machinery to post-transcriptionally repress the expression of their target mRNAs, by mRNA degradation, translational inhibition or transcript destabilization (Winter et al., 2009). Accumulating evidence has established the diverse modulatory roles of miRNAs in physiological processes and pathological conditions, ranging from growth and development to tumorigenesis (Johnson et al., 2005;

Kanellopoulou et al., 2005; Lu et al., 2005).

1.4.1 Roles of microRNAs in CD8+ T cell fate and function

Given their impact on health and development, it is unsurprising that miRNAs exert integral and effective regulation over the adaptive immune system (Baltimore D,

2008; Ebert PJ, 2009; Koralov SB, 2008 ; Li QJ, 2007; Rodriguez A, 2007; Thai TH,

2007 ; Xiao C, 2008). MiRNAs have been reported to modulate all aspects of T cell biology, including T cell development, activation and functional differentiation. Of particular interest, miRNA-mediated post-transcriptional gene regulation is intricately intertwined with the effector and memory differentiation of CD8+ T cells. This is evident 41

from the aberrant functional phenotype of CD8+ T cells lacking mature miRNAs: despite exhibiting accelerated proliferation and increased expression of the cytolytic degranulation machinery, activated Dicer-deficient CD8+ T cells have survival defects and traffic to peripheral tissues less efficiently (Trifari et al., 2013; Zhang and Bevan,

2010). MiRNAs are thus fundamental regulatory elements that fine-tune and shape CD8+

T cell responses.

In fact, miRNA expression in CD8+ T cells are highly dynamic, with naïve, effector and memory cells exhibiting differential miRNA profiles (Khan et al., 2013; Wu et al., 2007); this underscores the importance of synchronizing miRNA expression with discrete differentiation states of a CD8+ T cell. In this regard, antigen-induced activation and cytokine signals have been reported to coordinately elicit dramatic changes to the landscape of miRNA expression within CD8+ T cells, in turn regulating their fates and function (Sheppard et al., 2014; Trifari et al., 2013). For instance, the exposure of activated CD8+ T cells to highly-inflammatory environments suppresses miR-139 expression, permitting the up-regulation of perforin and Eomes for cytotoxic potency and efficient pathogen clearance (Trifari et al., 2013). In addition, TCR signaling strength determines the extent of miR-155 induction, which is crucial for both the effector function and memory formation of CD8+ T cells (Dudda et al., 2013). miR-155 not only promotes effector CD8+ T cell proliferation for efficient anti-viral and anti-tumor responses, but also supports effector cell survival for transition into long-lived memory cells (Dudda et al., 2013; Gracias et al., 2013). Moreover, the levels of the miRNA cluster, miR-17-92, induced by T cell activation have been shown to dictate the fate of

effector CD8+ T cells: while high miR-17-92 expression skews terminal effector differentiation to SLECs, low miR-17-92 levels favors the generation of MPECs (Khan et al., 2013). Manipulating specific miRNAs in CD8+ T cells therefore represents a means of fine-tuning the immune response and offers new therapeutic targets for immune- modulation.

1.4.2 Tools for targeting microRNAs

Our greater appreciation of how specific miRNA species contribute to disease pathogenesis has sparked the recent development of miRNA-based approaches for therapeutic intervention. Targeting miRNAs is advantageous, because it allows the properties of a cell to be altered without the need for complex transcriptional reprogramming. Moreover, miRNA-based therapy offers two advantages over conventional protein-target-based immune-modulation – it is more straightforward to engineer oligonucleotide-based anti-sense miRNA inhibitors, and advances in the field of chemical engineering have made in vivo delivery feasible (Kota J, 2009; Krützfeldt J,

2005; Obad S, 2011; Thum T, 2008). Two methods commonly used to functionally inhibit miRNAs will be discussed.

MiRNA loss-of-function can be achieved pharmacologically through the use of oligonucleotide inhibitors bearing sequence complementarity to the target miRNA.

Invariably designed to competitively bind and sequester endogenously-expressed miRNA away from their bona fide mRNA targets, these anti-sense miRNA inhibitors typically come in two main flavors: chemically-modified oligonucleotides and transgenic miRNA decoys. 43

Anti-sense oligonucleotides are chemically modified to enhance their binding affinities to their cognate miRNA (e.g. with locked nucleic acids (LNA), 2′-O-methyl (2′-

O-Me) or 2′-O-methoxyethyl (2′-MOE)), increase their nuclease resistance and biostability within the cell (e.g. with phosphorothioate backbone modifications) and to promote cellular uptake (Krutzfeldt et al., 2005; Obad et al., 2011; Orom et al., 2006).

Antagomirs and anti-miR-LNAs represent the two most potent forms of oligonucleotide- based pharmacological miRNA inhibitors. Usually comprised of 20-24 nucleotides with

2′-O-Me, phosphorothioate and cholesterol modifications, antagomirs are perfectly complementary to the entire length of the mature target miRNA (Krutzfeldt et al., 2005).

On the other hand, anti-miR-LNAs are phosphorothioated 8-mer sequences bearing perfect complementarity to only the seed region (nucleotides 2-7) of the target miRNA

(Obad et al., 2011; Orom et al., 2006). Clearly, the chemical modifications involved make the synthesis of these miRNA inhibitors a complicated process. Both antagomirs and anti-miR-LNAs have their pros and cons, making them comparably effective: the cholesterol modifications of antagomirs enables more efficient cellular uptake than naked anti-miR-LNAs, but the superior affinity of LNAs for their miRNA targets compensates for their potency at lower intracellular concentrations. While these pharmacological miRNA inhibitors can successfully and specifically silence their target miRNAs in vivo, their effects are short-lived and therefore require repeated delivery for permanent miRNA inhibition (Krutzfeldt et al., 2005; Obad et al., 2011). In our hands, we have observed a similar phenomenon: murine T cells actively exocytose these oligonucleotide-based miRNA inhibitors, especially following T cell activation (data not shown). Moreover,

when intravenously administered, these agents are promiscuously taken up by multiple cell types and tissues throughout the body, negating the targeted cell-specific inhibition of miRNAs (Krutzfeldt et al., 2005; Obad et al., 2011). As a result, antagomirs and anti- miR-LNAs are predominantly used to probe miRNA function in short-term studies in vitro.

Alternatively, transgenic miRNA decoys can be employed to achieve long-lasting miRNA suppression. In contrast to antagomirs, miRNA decoy sequences are most effective when they are designed with only partial complementary to their target miRNAs: miRNA decoys bear complementarity to the seed region, but are mismatched with nucleotides 9-12 of their target miRNAs. The imperfect pairing in the decoy:miRNA hybrid results in a bulge, thereby preventing RISC-mediated cleavage of the decoy and enhancing its stability (Ebert et al., 2007). Tandem repeats of the miRNA decoy sequence are inserted downstream of a strong promoter to enhance target binding affinity and ensure high copy numbers, facilitating maximal sequestration of the target miRNA (Ebert et al., 2007). When incorporated into lentiviral or retroviral vectors, stable expression of miRNA decoys in cells can easily be achieved by gene transduction and chromatin integration, permitting long-lasting inhibition of target miRNAs in genetically-modified cells (Ebert and Sharp, 2010). The decoy-expressing cell type of interest can then be selected and adoptively transferred for long-term in vivo studies or therapeutic applications. These well-established methods to functionally inhibit specific miRNAs of interest therefore offer a readily available toolset for T cell immune-modulation.

1.5 Targeting auto-reactive effector T cells for the treatment of inflammatory diseases

Autoimmunity is a misdirected immune response that occurs when the immune system goes awry and attacks the body itself. More than 80 human diseases currently are classified as autoimmune, including multiple sclerosis, inflammatory bowel diseases, type 1 diabetes mellitus, rheumatoid arthritis, and systemic lupus erythematosus. Since they affect up to 8% of the US population (The Autoimmune Diseases Coordinating

Committee, 2005) and often attack young adults, especially women, their social and economic impact is enormous (Cooper and Stroehla, 2003). In many cases, autoimmunity arises from a breakdown in T cell tolerance. Pathologically, this failure in tolerance results in chronic lymphocyte activation, sustained leukocyte and lymphocyte tissue infiltration, massive production of inflammatory cytokines, and secretion of autoantibodies (Goodnow et al., 2005). The pathogenicity of auto-reactive T cells in these processes has spurred a keen interest in the development of immune-suppressive agents targeting the various aspects of T cell activation and effector function.

1.5.1 The unmet need for tolerable T cell-specific immune- suppressants

Immunosuppressive drugs are classical therapies to treat a wide range of autoimmune diseases (Stepkowski, 2000). In the last decade or so, a few new immunosuppression medications have been approved, increasing the number of options available to treat autoimmune diseases. However, currently available immunosuppressive drugs possess serious side effects (Stepkowski, 2000).

Antiproliferative immunosuppressive drugs aimed at inhibiting the expansion of pathogenic effector T cells, such as methotrexate and cyclophosphamide, are limited in that they have non-specific effects on various types of proliferating cells. They cause serious non-specific bone marrow suppression, impair host resistance, and increase the incidence of infections. They also have a slow onset of action and a moderate efficacy that declines after several years of treatment.

In comparison with these non-selective antiproliferative agents, cyclosporine A

(CsA), FK506, and rapamycin act more selectively on different stages of the T- and B- lymphocyte activation cycles. However, even these more selective immunosuppressive drugs possess serious side effects, including acute neurological toxicity, chronic nephrotoxicity, biphasic effects to bone structure, and hypertriglyceridemia. Antibody- based immune-suppressive biologics, such as integrin blockers (e.g. the anti-α4-integrin

Natalizumab and the anti-αL-integrin Efalizumab) and anti-TNFα, have faster onset of action and higher specificity than the existing small molecule-based therapies (Balague et al., 2009; Sathish et al., 2013). However, they are expensive and can trigger severe side effects. For instance, the treatment of multiple sclerosis and Crohn’s disease patients with integrin blockers has been reported to compromise immune-surveillance mechanisms, due to reduced T cell trafficking to the central nervous system (CNS) (Sorensen et al.,

2012). This leaves patients susceptible to severe central nervous system viral infections that culminate in progressive multifocal leukoencephalopathy (Sorensen et al., 2012).

Anti-TNFα therapy for rheumatoid arthritis, too, resulted in the reactivation of latent infections and an increased risk of opportunistic bacterial infections(Balague et al., 2009).

There therefore remains an unmet need for immune-suppressive drugs with improved safety profiles. This calls for the continued discovery and characterization of novel immunosuppressive agents capable of targeting specific pathways in key pathogenic cell types. In this regard, auto-reactive T cells may offer a myriad of potential drug targets, due to the unique signaling intermediates and cellular processes involved in

T cell activation, survival and metabolism.

2. Materials and Methods

2.1 Mice

pMel-1 mice carrying a transgenic T cell receptor specific for the B16 melanoma antigen gp100 (C57BL/6-Tg(TcraTcrb)8Rest/J ), OT-I mice (C57BL/6-

Tg(TcraTcrb)1100Mjb/J) and C57BL/6J mice were purchased from the Jackson

Laboratory. The LLO118 TCR transgenic mice specific for listeriolysin LLO190-205 were made from T-cell hybridomas generated from Listeria monocytogenes-infected mice (Weber et al., 2012). All mice were housed under pathogen-free conditions, and used between 6 to 10 weeks of age for experimental procedures.

2.2 Cell culture

T cells and EL4 thymoma cells were cultured in RPMI1640 media supplemented with 10% FBS, 100 U/ml penicillin, 100 U/ml streptomycin, 2 mM L-glutamine, 1 mM sodium pyruvate, 0.1 mM non-essential amino acids and 50 μM 2-mercaptoethanol

(henceforth referred to as complete RPMI) in a humidified 37°C incubator with 7% CO2.

The murine 3T3 fibroblast cell line was cultured in DMEM media supplemented with 10% FBS, 100 U/mL penicillin, 100 U/mL streptomycin, 2 mM L-glutamine, 1 mM sodium pyruvate, and 0.1 mM non-essential amino acids.

BMDCs were generated as previously described (Yang and Baltimore, 2005).

Briefly, total bone marrow cells were harvested from the femurs of B6 mice, and cultured in complete RPMI supplemented with 1:30 J558L conditioned media. The J558L cell line, a kind gift from Dr. David Baltimore (Caltech), had been stably transfected with the

murine GM-CSF cDNA and its cell culture supernatant was used as a source of GM-CSF

(Zal et al., 1994). From days 3 through 9, cells were fed every other day. Non-adherent cells were collected on day 9 and treated with 1 μg/ml LPS (Sigma) overnight. On day

10, non-adherent mature DCs were collected, and subjected to density gradient centrifugation over Histopaque (Sigma). The viable mature DCs isolated were washed three times in complete RPMI before co-culture with naïve T cells.

Lymph nodes harvested from mice were teased apart into a single-cell suspension and passed through a 70 μm cell strainer. Naïve T cells were isolated from lymph nodes and/or spleens using the Dynal® mouse CD8 negative isolation kit (Invitrogen) according to the manufacturer’s instructions. For T cell priming by antigen presenting cells, naïve pMel-1 CD8+ T cells were co-cultured with either mature DCs or sorted immature B220+ splenic B cells at a 1:1 ratio, in the presence of 5 μM hgp100 25-33 peptide. For T cell activation by antibodies, naïve pMel-1 CD8+ T cells were seeded onto plate-bound αCD3 and αCD28 antibodies (5 μg/ml each, unless otherwise indicated;

Biolegend). Lymphocytes were cultured in complete RPMI1640 media in a humidified

37 °C incubator with 7% CO2.

2.3 miRNA expression profiling and miRNA qPCR

Naïve pMel-1 CTLs were primed by mature DCs or splenic B cells in vitro, as described above. After 3 days of priming, TCRβ+ pMel-1 CTLs were sorted and lysed using the RNAqueous Micro-kit (for samples containing 10 5 – 0.5×10 6 cells) or the mirVana miRNA Isolation kit (for samples containing ≥ 0.5×10 6 cells) (both from

Ambion) according to the manufacturer’s instructions. 50

To quantify the expression of mature miRNA expression, E. coli poly A polymerase (Epicentre) was first used to generate polyadenylated tails at the 3’-end of all

RNA molecules. After annealing oligo-dT primers, cDNA was synthesized using the qScript™ Flex cDNA synthesis kit (Quanta Biosciences) as per the manufacturer’s instructions for gene specific priming, with one modification: a universal tag that would extend from the 3’-end of cDNA molecules was added during reverse transcription. With the addition of this universal tag, individual miRNAs were detected with miRNA-specific forward primers and a reverse universal primer mix. A SYBR Green-based real-time

PCR method was used to quantify the relative expression of mature miRNAs. In the miRNA expression profiling array, a total of 355 mature miRNAs were evaluated in DC- and B cell-primed CTLs (n=3 independent experiments). miRNA expression was normalized using by geometric mean-based global normalization using the Realtime

StatMiner (Integromics) analysis software. Differential miRNA expression was determined by paired t-test, with significance level set at 0.05. The complete set of miRNA expression profiling data is available on the NCBI Gene Expression Omnibus database under the accession number GSE60884.

2.4 Target prediction and luciferase reporter assays

Candidate targets of miR-23a were derived from the integrated miRNA target prediction resource miRecords (http://mirecords.biolead.org/). The full-length 3’UTRs of mouse prdm1, eomes and tbx21 were amplified from a 3’RACE-ready cDNA library generated from total mouse T cell RNA, and cloned into the pmirGLO dual-luciferase vector (Promega) downstream of firefly luciferase. Each dual-luciferase reporter vector, 51

together with a Mock or miR-23a overexpression vector, was co-transfected into Jurkat T cells using the Amaxa Cell Line Nucleofactor kit (Lonza). 48 hours post-transfection, cells were lysed and luciferase reporter activities were determined in a dual-luciferase reporter assay (Promega).

2.5 mRNA and pri-miRNA qPCR

Total RNA from cells was isolated using the RNAqueous Micro-kit (for samples containing 10 5 – 0.5×10 6 cells) or the mirVana miRNA Isolation kit (for samples containing ≥ 0.5×10 6 cells) (both from Ambion) according to the manufacturer’s instructions. cDNA was reverse-transcribed from total RNA using a mixed priming strategy (oligo-dT and random primers) with the qScript™ Flex cDNA synthesis kit

(Quanta Biosciences) as per the manufacturer’s instructions. A SYBR Green-based realtime PCR method was used to quantify the relative expression of mRNAs and pri- miRNAs.

2.6 Lymphocyte isolation and miR-23a quantification from lung cancer patients

Pleural effusion samples were collected from newly-diagnosed lung cancer patients with malignant pleural effusion (MPE). Patients included in this study neither underwent any invasive procedures directed into the pleural cavity, nor suffered chest trauma within the 3 months prior to hospitalization. At the time of sample collection, none of the patients had received any anticancer therapy, corticosteroids, or other nonsteroid anti-inflammatory drugs. Pleural fluid samples were collected in heparin- treated tubes from each subject, using a standard thoracocentesis technique. Twenty

milliliters of peripheral blood was drawn simultaneously. MPE and peripheral blood lymphocytes were isolated by density centrifugation using human lymphocyte separation medium (TBD, Tianjin, China) according to the manufacturer’s instructions.

Total RNA extraction, as well as miRNA and mRNA qPCR, was performed as described above. For candidate endogenous controls, hsa-RNY3, hsa-U6 and hsa-U1 were included for miRNA qPCR; while 18S RNA, RPLP0 and RPL13A were included for mRNA qPCR. Using the Realtime StatMiner (Integromics) analysis software,

Genorm analysis was performed and the mean Ct-values of RNY3 and U6 were chosen as internal controls for miRNA Ct normalization; while the mean Ct-values of 18S RNA and RPLP0 were chosen as internal controls for mRNA Ct normalization. For each patient, ddCt of miR-23a and prdm1 were then calculated from the difference between dCt-values in TIL and PBMC samples, and transformed into a fold change. The relative expression of miR-23a and prdm1 in PBMC of each patient was arbitrarily set to 1.0. ddCt of IFNγ mRNA was calculated from the difference between the dCt-value of each sample and the TIL sample with the lowest dCt-value, and transformed into a fold change.

2.7 miR-23a decoy construct design

The miR-23a decoy vector is a bicistronic retroviral backbone that encodes two independent expression cassettes: the 5’-LTR drives the expression of the selectable marker (either iRFP (Filonov et al., 2011) or puromycin-resistance), whereas the PGK promoter drives the expression of a GFP decoy/reporter. The GFP reporter contains an insertion of eight tandem miR-23a binding sites 53

(GGAAATCCCTG cg AATGTGAT cgtt GGAAATCCCT cc CAATGTGAT actc GGAA

ATCCC ac GCAATGTGAT gtac GGAAATCCC acc CAATGTGAT ccga GGAAATCC

CT ccg AATGTGAT acgc GGAAATCCCT cc CAATGTGAT ccta GGAAATCCC acc CA

ATGTGAT agct GGAAATCC gac GCAATGTGAT ) in its 3’UTR to monitor the sponge effect of decoy targeting sites. Between these two cassettes, we additionally inserted an insulator sequence comprising two tandem repeats of the chicken β-globin

FII/FIII∆spacer insulator fragment (FII/FIII∆spacer: aggcgcgcccccagggatgtaattacgtccctcccccgctagggggccggccagcaccggtccggcgctccccccgcatcc ccgagccggggcgcgcct) (Bell et al., 1999) to maximize their independent expression. As a control, a Mock decoy vector lacking miR-23a binding sites in its GFP 3’UTR was also generated.

2.8 Retroviral transduction

2.8.1 miR-23a overexpression and decoy vectors

On Day 0, cells from the lymph nodes of pMel-1 and OT-1 mice were harvested and seeded into 24-well plates. Non-T cells in the lymph nodes served as antigen- presenting cells, and CTLs were primed in vitro by the addition of 5 μM hpg100 25-33 or

OVA 257-264 , respectively. On Day 1, 50 U/ml mIL-2 was added. 6 hours later, cells were spin infected with retroviral supernatants at 1250 g for 90 minutes at 37°C. CTLs from days 4 through 6 were used in experiments.

2.8.2 EGFRvIII-CAR vector

On Day 0, cells from the lymph nodes of ER-Cre+ miR-23af/f pMel-1 mice were harvested, seeded into 24-well plates and activated with 2.5 μg/ml concanavalin A and 50 54

U/ml mIL-2. Cells were simultaneously treated with 200 nM 4-hydroxytamoxifen to induce miR-23a deletion, or with 100% ethanol as vehicle control. On day 2, cells were spin infected with retroviral supernatants at 1250 g for 90 minutes at 37°C. CD8+ T cells were restimulated on Day 4 for functional assays.

2.9 Western blot

For Western blot analysis of protein expression, CTLs cultured for the indicated times were lysed in RIPA buffer containing protease inhibitor (Roche) and phosphatase inhibitor cocktails 1 and 2 (Sigma). Samples were run on 10% polyacrylamide gels

(BioRad) and transferred onto PVDF membranes. Proteins of interest were probed with the following primary antibodies: mouse-α-Blimp-1 (Biolegend), rabbit-α-pSmad2, rabbit-α-Smad2/3 (both from Cell Signaling) and goat-α-βactin (Sigma). α-mouse-

AlexaFluor680, α-rabbit-AlexaFluor680 and α-goat-AlexaFluor680 (all from Invitrogen) were used as secondary antibodies. Florescence intensity was measured on an Odyssey imaging system (LI-COR Biosciences).

2.10 In vitro cytotoxicity assays

EL4 cells were pulsed with 10 μM hgp100 25-33 or 10 μM OVA 257-264 overnight.

Peptide-pulsed and unpulsed EL4 cells were labeled with 5 μM and 0.5 μM Cell Tracker

Orange (Invitrogen), respectively, and mixed in a 1:1 ratio before co-culture with CTLs.

Following 6 days of in vitro culture, viable pMel-1 CTLs were isolated by density gradient centrifugation over Histopaque (Sigma), and co-cultured with labeled EL4 target cells at an effector:target ratio of 5:1 for 6 hours in a humidified 37°C incubator. iRFP+ and/or GFP+ OT-1 CTLs expressing the MSCV-iRFP-2Xins-mG-Mock or MSCV-iRFP- 55

2Xins-mG-miR-23a Decoy vectors were sorted 48 hours after retroviral transduction, and co-cultured with EL4 target cells at the indicated effector:target ratios for 6 hours in a humidified 37°C incubator. Following 6 hours of co-culture, samples were harvested and stained with the Live/Dead® Violet viability kit (Invitrogen) and α-CD8α -FITC

(Biolegend). CountBright™ Absolute counting beads (Invitrogen) were added to samples before acquisition on the FACSCanto II flow cytometer (BD), and data were analyzed using the FlowJo software.

2.11 CD4+ T cell differentiation assays

Naïve LLO118 lymphocytes were activated with 10 μM LLO 190-205 peptides and cultured under the respective Th-differentiating conditions for a total of 6 days. For Th0 differentiation, no additional cytokines or antibodies were added. For Th1 differentiation,

50 U/ml IL-2 (Biolegend), 50 ng/ml IL-12 (R&D Systems), and 10 μg/mL anti-IL-4

(Clone 11B11; Biolegend) were added. For Th17 differentiation, 50 ng/ml IL-6

(Biolegend), 4 ng/mL TGFβ (Biolegend), 10 μg/mL anti-IL-4 (Clone 11B11; Biolegend), and 10 μg/mL anti-IFNγ (Clone XMG1.2; Biolegend) were added. During the final 2 days of culture, the indicated amounts of drugs or DMSO vehicle control were added.

After 2 days of drug treatment, CD4+ LLO118 T cells were subjected to a 4h restimulation and intracellular staining as described in Chapter 2.13.

2.12 T cell restimulation

For cytokine staining, CD4+ LLO118 murine T cells, pMel-1 CD8+ murine T cells and human lymphocytes were restimulated for 4 hours with 0.9 nM PDBu (Sigma)

and 0.5 μg/ml ionomycin (Sigma), in the presence of 5 μg/ml Brefeldin A (eBioscience) and 2 μM Monensin A (eBioscience).

2 days after EGFRvIII-CAR retroviral transduction, viable wildtype and miR-23a- deficient pMel-1 T cells were enriched by Ficoll (Histopaque, Sigma) density gradient centrifugation. Viable cells were then restimulated for 24 hours on 1 μg/ml plate-bound

PepvIII peptide, in the absence or presence of the indicated concentrations of TGFβ

(Biolegend). During the final 4 hours, 5 μg/ml Brefeldin A (eBioscience) and 2 μM

Monensin A (eBioscience) were added prior to staining.

2.13 Intracellular staining and flow cytometry

Viable cells were stained using the Live/Dead® Violet viability kit (Invitrogen) as per the manufacturer’s instructions prior to intracellular staining. Intracellular staining for cytokines, granzyme B and transcription factors were performed by fixing cells in 2% paraformaldehyde, followed by membrane permeabilization in 0.1% saponin. Fc receptors were blocked before incubation with the following panels of staining antibodies: (i) α-CD8α-PECy7, α-T-bet-PE, α-granzyme B-AlexaFluor647 and α-IFNγ-

PE (all from Biolegend); (ii) α-CD8α-PECy7, α-Eomes- AlexaFluor647 and α-granzyme

B-PE (all from eBioscience); and (iii) α-CD4-Pacific Blue, α-IL-2-PE, α-IFNγ-PECy7 and α-IL-17A-APC (all from Biolegend). Samples were acquired on the FACSCanto II flow cytometer (BD), and data were analyzed using the FlowJo software.

2.14 Proliferation and cell viability assays

Naïve pMel-1 and LLO118 lymphocytes were stained with 10 μM carboxyfluorescein diacetate succinimidyl ester (CFSE; Invitrogen) at a density of 5 · 10 6 cells/ml for 10 minutes at 37 °C. Cells were then washed twice with cold complete RPMI and activated with 5 μM gp100 25-33 or 10 μM LLO 190-205 peptides, together with 100 nM of drugs or DMSO vehicle control. With each successive cell division, CFSE staining intensity decreases by half, enabling the extent of cell proliferation to be determined.

After 48 hours of culture, CFSE-labelled CD8+ pMel-1 and CD4+ LLO118 T cells were assessed for cell death using the Live/Dead Violet viability kit (Invitrogen) according to manufacturer’s instructions. Samples were acquired on the FACSCanto II flow cytometer

(BD), and data were analyzed using the FlowJo software.

2.15 Labeling and functional analysis of mitochondria

Naïve LLO CD4+ T cells were cultured under Th1- or Th17-polarizing conditions for 4 days, and treated with various doses of subglutinol A for an additional 2 days. To interrogate immature BMDC (imBMDC) mitochondrial integrity, non-adherent cells from BMDC cultures were collected on day 9, and treated for 24 hours with 100 nM of drugs or DMSO vehicle control. To interrogate mature BMDC (mBMDC) mitochondrial integrity, non-adherent cells from BMDC cultures collected on day 9 were matured with

1 μg/mL lipopolysaccharide (Sigma) for 24 hours, with concurrent treatment of 100 nM drugs or DMSO vehicle control. Cells were then labeled with 100 nM MitoTracker Deep

Red FM (Invitrogen) and 200 nM MitoTracker Orange CMTMRos (Invitrogen) for 15 minutes at 37 °C, and washed. Following mitochondria labeling, cells were stained with 58

the Live/Dead Violet viability kit (Invitrogen), together with α-CD4-PECy7 (Biolegend) to label T cells or α-CD11c-PECy5 (Biolegend) to label dendritic cells. Samples were acquired on the FACSCanto II flow cytometer (BD), and data were analyzed using the

FlowJo software.

2.16 Delayed-type hypersensitivity induction and treatment

To induce primary sensitization, each C57BL/6J mouse was immunized subcutaneously with 100 μg keyhole limpet hemocyanin protein (KLH; Sigma) emulsified in 100 μL complete Freund’s adjuvant (Sigma) on Day 0. On Day 7 (one week post-sensitization), the thickness of both footpads was measured prior to KLH rechallenge. The left footpad of each mouse was immediately rechallenged with 50 μg

KLH, together with 16 nmol (0.273 mg/kg) of subglutinol A, 16 nmol (0.769 mg/kg) of

CsA, or DMSO (5 mice per group), solubilized in 25 μL PBS; as a control for tissue damage or inflammation caused by injection, the right footpad of each mouse was injected with an equal volume of PBS vehicle control. On Day 8, a second treatment of

16 nmol (0.273mg/kg) of subglutinol A, 16 nmol (0.769 mg/kg) of CsA, or DMSO solubilized in 25 μL PBS was administered into the left footpads, while an equal volume of PBS was injected into the right footpads. On Day 9 (48 hours post-rechallenge), footpad thicknesses were measured. Footpad swelling was calculated from the difference in footpad thicknesses between that of Day 7 (before rechallenge) and Day 9 (after rechallenge). Mice were sacrificed and their footpads excised, inflated with 4% paraformaldehyde in PBS, fixed overnight at room temperature, placed in 70% ethanol and embedded in paraffin, prior to hemotoxylin & eosin staining. 59

2.17 In vivo tumor models

2.17.1 Subcutaneous tumor models

The B16/F10 melanoma and LLC-OVA lung cancer cell lines were kind gifts from Dr. Thomas Tedder (Duke University) and Dr. Eckhard Podack (University of

Miami), respectively. Tumor cells were harvested by trypsinization, and cell viabilities of

>95% were confirmed by trypan blue exclusion. To study the in vivo anti-tumor effects of miR-23a-overexpressing pMel-1 CTLs, 0.2×10 6 B16/F10 cells in 200 μl PBS were inoculated subcutaneously into the shaved right lateral flanks of B6 recipient mice on

Day -3. 3 days after tumor inoculation, each recipient mouse received an intravenous adoptive transfer of either 0.6×10 6 sorted GFP+7AAD- Mock or miR-23a overexpressing pMel-1 CTLs in 200 μl PBS on Day 0. Control mice not treated with CTLs received intravenous injections of 200 μl PBS alone. To study the in vivo therapeutic potential of miR-23a-inhibited pMel-1 and OT-I CTLs, 1×10 6 LLC-OVA cells in 200 μl PBS were inoculated subcutaneously into the shaved right lateral flanks of B6 recipient mice on

Day -7. On Days 0 and 5, each recipient mouse received two intratumoral injections of either 0.2×10 6 sorted Mock or miR-23a Decoy-expressing CTLs in 50 μl PBS. Tumor progression was monitored closely, and tumor volumes were calculated using the equation: V = 4π (L1 x L2)/3, where V = volume (mm3), L1 = longest radius (mm), L2 = shortest radius (mm). Mice were sacrificed at the experimental end-points, and their spleens, draining lymph nodes and tumors harvested. Tumors were digested using the

Papain Dissociation System (Worthington Biochemical) to liberate tumor-infiltrating

cells. Effector functions of the transferred pMel-1 or OT-I CTLs were then analyzed by flow cytometry.

For granzyme B inhibition studies, transduced pMel-1 CTLs were pre-treated with 12.5 μM of the granzyme B inhibitor zAAD-CMK (Enzo Life Sciences) or DMSO vehicle control for 48 hours in vitro, prior to intratumoral injection. 3 days after the transfer of granzyme B-inhibited CTLs, an additional 10 μg zAAD-CMK or DMSO (both solubilized in PBS) was intratumorally administered to sustain granzyme B inhibition in vivo.

2.17.2 Intracerebral glioblastoma model

On Day -14, 2×10 4 KR158B glioma cells stably expressing luciferase and

EGFRvIII (KLucvIII) were injected intracerebrally into C57BL/6 mice, and were allowed two weeks to form established orthotopic tumors. On Day 0, 2×10 6 EGFRvIII-CAR transduced wildtype pMel-1 T cells, 2×10 6 EGFRvIII-CAR transduced miR-23a-deficient pMel-1 T cells, or saline was delivered intracerebrally. On Day 10, tumor burdens were assessed by bioluminescence imaging, after which mice were sacrificed for the collection of brain tumor tissues and spleens.

2.18 T cell repertoire library construction

Brain tissues were immediately snap-frozen in liquid nitrogen and homogenized in Trizol. After homogenization and red blood cells lysis, 1×10 6 total splenocytes were lysed in 1 ml Trizol. Total RNA from brain tissues and splenocytes were purified by phenol-chloroform extraction, and isolated using the Direct-zol RNA miniprep kit (Zymo

Research) according to the manufacturer’s instructions. First-strand cDNA synthesis was 61

carried out using a murine TCRB constant region-specific primer (5’-ACT GTG GAC

CTC CTT GCC A-3’) with the qScript™ Flex cDNA synthesis kit (Quanta Biosciences) as per the manufacturer’s instructions. A multiplex PCR system was utilized to amplify the CDR3 region of rearranged TCRB loci. A set of forward primers, each specific to one or a set of functional TCR V beta segments, and a reverse primer specific to the constant region of TCRB , were used to generate amplicons spanning the entire CDR3 region. PCR products were separated on 3% agarose gels (Sigma-Aldrich), and bands ~220-240 bp in size were then excised and purified using the QIAquick Gel Extraction kit (QIAGEN).

Purified products were sequenced using the Ion Torrent PGM platform (Life

Technologies).

2.19 Statistical analysis

The two-tailed unpaired or paired Student’s t-tests were applied for the comparison of two means. For multiple comparisons, the one-way or two-way ANOVA with Bonferroni post-test were performed as indicated. To assess the correlation between miR-23a and mRNA expression in CD8+ T cell samples from lung cancer patients, the

Pearson’s correlation coefficient was calculated. P-values <0.05 were considered statistically significant.

3. Mechanistic significance of targeting miR-23a in CD8+ T cells for ACT

The contents in this chapter were originally published in the Journal of Clinical

Investigation. Targeting miR-23a in CD8+ cytotoxic T lymphocytes prevents tumor- dependent immunosuppression. Regina Lin, Ling Chen, Gang Chen, Chunyan Hu, Shan

Jiang, Jose Sevilla, Ying Wa3, John H. Sampson, Bo Zhu and Qi-Jing Li.

3.1 Introduction

Owing to their unique abilities for specific tumor antigen recognition and prompt execution, CD8+ cytotoxic T lymphocytes (CTLs) represent the primary leukocyte population employed for adoptive cell transfer therapy (ACT) in cancer treatment.

Conventional CTL-based tumor immunotherapy relies on isolation, followed by extensive ex vivo expansion of tumor-infiltrating CTLs in the presence of copious amounts of growth factors (e.g. IL-2) in vitro, followed by autologous reinfusion into the patient (Rosenberg, 2001; Rosenberg SA, 2008). Recent advances in CTL engineering have allowed the enforced expression of high affinity and tumor-specific T cell receptors

(TCRs) or chimeric antigen receptors (CARs), thereby mitigating part of the difficulties in CTL isolation and expansion, as well as the efficacy of tumor antigen targeting

(Chhabra, 2011; Park et al., 2011; Rosenberg SA, 2008). Paradoxically, although these strategies are capable of generating tumor-specific CTLs exhibiting potent cytotoxic profiles in vitro , clinical success of ACT using ex vivo IL-2-conditioned and TCR- 63

redirected CTLs has been partial at best – the majority of patients fail to respond with complete tumor regressions (Pilon-Thomas S, 2012; Robbins PF, 2011; Rosenberg SA,

2011).

This apparent discrepancy between the in vitro and in vivo functionality of CTLs in ACT is largely attributed to the presence of immunosuppressive barriers within the tumor microenvironment, which are co-opted by tumors to evade the host immune system (Fourcade J, 2010; Prado-Garcia et al., 2012; Woo SR, 2012). Of these, TGFβ is a key pathogenic cytokine that is secreted and up-regulated by a wide variety of tumors, including melanoma and lung cancer (Bennicelli JL, 1993; De Jaeger K, 2004; Derynck

R, 1985; Polak ME, 2007). Studies in melanoma and lung cancer patients have found plasma TGFβ levels to be a negative prognostic indicator of tumor progression – high levels of circulating TGFβ is associated with increased metastasis and relapse rates, as well as decreased overall patient survival (Kong F, 1999; Krasagakis K, 1998; Zhao L,

2010 ). Moreover, local expression of TGFβ is further elevated within metastatic melanoma lesions, as compared their primary tumors (Van Belle P, 1996). TGFβ promotes tumor outgrowth and metastasis in various avenues, a critical one of which is to hamper productive anti-tumor immune responses. Specifically, TGFβ-induced Smad signaling in both naïve and full-fledged effector CTLs represses their expression of key cytotoxic molecules, including granzyme B and IFNγ, resulting in CTL dysfunction and impaired tumor rejection (Ahmadzadeh M, 2005; Thomas DA, 2005). Therefore, engineering tumor-specific CTLs to overcome TGFβ-mediated immune-suppression and

preserve their cytotoxicity within the tumor microenvironment remains one of the holy grails in the field of cancer intervention, and continues to be an area of intensive research.

CTL function and cytotoxicity are governed by several key transcription regulators, including T-bet, Eomes and Blimp-1. In effector CTLs, T-bet and Eomes are compensatory and essential transcriptional factors enforcing a Type 1 program that instructs their differentiation into highly-potent killer CTLs – T-bet and Eomes drive the expression of Type 1 cytotoxic effector molecules (e.g. granzyme B, perforin and IFNγ) for the eradication of malignant cells, while simultaneously repressing the acquisition of an unproductive Type 17 program that targets extracellular pathogens (Cruz-Guilloty F,

2009; Intlekofer AM, 2008; Pearce EL, 2003; Sullivan BM, 2003). Unsurprisingly, the absence of T-bet and Eomes in CD8+ T cells severely compromises their cytotoxic capacity and anti-tumor responses (Intlekofer AM, 2008; Zhu Y, 2010 ). Likewise, the transcriptional regulator Blimp-1 is also essential for the acquisition of CTL effector function. Blimp-1 represses the quiescent transcriptional program characteristic of memory CTLs, thus promoting CTL effector differentiation (Kallies A, 2009;

Rutishauser RL, 2009; Shin H, 2009). Of note, Blimp-1-deficient effector CTLs have impaired cytotoxicity, and show reduced expression of multiple CTL effector molecules, including granzymes, perforin and T-bet (Kallies A, 2009; Rutishauser RL, 2009).

To enhance the efficacy of current tumor immunotherapy, we became interested in a novel microRNA (miRNA)-based approach to augment the cytotoxic capacity of tumor-specific CTLs ex vivo. miRNAs are a group of small non-coding RNAs that have emerged as key regulators of gene expression in plants and animals (Carrington JC,

2003). Importantly, mounting evidence indicates that miRNAs are integral and effective regulatory elements of the adaptive immune system (Ebert PJ, 2009; Jiang et al., 2011; Li

QJ, 2007; Rodriguez A, 2007; Xiao C, 2008), making the manipulation of miRNA levels in CTLs an attractive means of enhancing anti-tumor adaptive responses. In addition, miRNA-based therapy offers two advantages over conventional protein-target-based immune modulation – it is far more straightforward to engineer anti-sense miRNA inhibitors, and advances in the field of chemical engineering have made in vivo delivery feasible (Kota J, 2009; Krützfeldt J, 2005; Obad S, 2011; Thum T, 2008). Unfortunately, little is known of the “master miRNAs” that modulate effector CTL function.

To address this knowledge gap, we compared the miRNA expression profiles of poorly- and highly-cytotoxic CTLs generated under different priming conditions, and identified miR-23a as a key inhibitor of anti-tumor CTL effector responses in vitro and in vivo. We demonstrate that miR-23a down-regulates its mRNA target Blimp-1, and simultaneously inhibits the expression of multiple key CTL effector molecules and transcriptional regulators. Additionally, we establish cMyc and tumor-associated TGFβ as key determinants of miR-23a abundance in effector CTLs.

3.2 Results

3.2.1 Identification of miR-23a expression as a negative correlate of cytotoxicity of effector CTLs

To screen for key miRNA regulators of CTL effector responses, we employed different in vitro systems that are well-known for generating CTLs with different killing capacities. Naïve murine pMel-1 CTLs were primed in vitro with either mature bone

marrow-derived dendritic cells (DCs; Figure 1, A and B), or splenic B cells pulsed with the melanoma tumor-associated antigen, hgp100 (Overwijk et al., 2003). pMel-1 CTLs expanded with peptide-loaded syngeneic B cells displayed very poor cytotoxic capacity at an effector:target (E:T) ratio of 5:1; in contrast, CTLs expanded with DCs exhibited 5- fold higher cytotoxic potency (Figure 2A). Accordingly, we found that DC-primed CTLs expressed higher levels of key cytotoxic mediators (granzyme B and IFNγ), and upstream master regulators (T-bet, Eomes and Blimp-1) (Figure 2B). Killing deficiencies observed in B cell priming were previously reported to result from impaired granzyme B expression and increased activation-induced T cell death (AICD) (Castiglioni et al., 2005;

Fuchs and Matzinger, 1992), which could be overcome by IL-15 and IL-21 (Novy et al.,

2011; Zeng et al., 2005). Consistent with earlier reports (Castiglioni et al., 2005; Fuchs and Matzinger, 1992; Novy et al., 2011; Zeng et al., 2005), the addition of exogenous IL-

15 and IL-21 during B cell-priming partially rescued granzyme B expression (Figure 3A) and AICD (Figure 3B) in CTLs. However, when challenged with a high ratio of antigen- pulsed target cells (E:T = 5:1), these cytokines enhanced their in vitro cytotoxicity insignificantly (Figure 3C). Qualitative differences in CTL cytotoxicity induced by DCs and B cells indicate that these two priming conditions elicit distinct cytotoxic transcriptomes in CTLs; therefore, as unphysiological as it may be, our in vitro priming system provides us with a comparative platform for discovering master regulator(s) of cytotoxicity.

As miRNAs can simultaneously regulate the expression of multiple genes post- transcriptionally (Bartel, 2009; Li et al., 2007), miRNA-based immunotherapy holds the

potential to bypass the need for complex transcriptional reprogramming of effector CTLs.

We therefore sought to identify miRNAs that modulate cytotoxicity using our in vitro priming system. After 3 days of priming with either DCs or B cells, we isolated the differentially-primed CTLs for miRNA expression profiling (Zhang et al., 2013b).

Among the 350 miRNAs screened, 18 were significantly differentially-expressed (Figure

4): 13 miRNAs were down-regulated and 5 were up-regulated in DC-primed CTLs

(Figure 5 and Table 1). To determine whether these miRNA candidates directly impacted

CTL cytotoxicity, we assessed granzyme B expression in pMel-1 CTLs overexpressing either the respective miRNAs, or a Mock-GFP control vector. Only miR-23a was able to inhibit both granzyme B and T-bet expression in CTLs (Figure 6 and data not shown).

Interestingly, although miR-23b – a paralog of miR-23a – was similarly suppressed in

DC-primed CTLs (Figure 5 and Table 1), miR-23b overexpression did not affect granzyme B levels (Figure 7). Further validation experiments corroborated that miR-23a expression in CTLs was dramatically suppressed during DC-priming (Figure 8). While miR-23a did not affect CTL proliferation (Figure 9A) and AICD (Figure 9B), overexpression of miR-23a (~3.4±1.0-folds increase from (3.5±0.6)×10 4 copies/cell;

Figure 10) blunted the expression of multiple key CTL effector molecules and transcription factors in vitro (Figure 6). These data suggest that miR-23a may negatively regulate CTL cytotoxicity.

Figure 1: Phenotype of splenic B cells and LPS-matured BMDCs (mDCs) used for naïve CTL priming in vitro.

(A) CD11c expression on LPS-matured BMDCs. (B) Surface expression of CD80 and CD86 on splenic B cells and mDCs. Numbers within histograms indicate mean fluorescence intensity.

Figure 2: Priming by splenic B cells and mDCs generate CTLs with different cytotoxic profiles.

pMel-1 CTLs primed with splenic B cells or LPS-matured bone marrow-derived DCs for 4 to 6 days were assessed for (A) in vitro cytotoxicity at an E:T ratio of 5:1 and (B) expression of CTL effector molecules. Histograms are representative of n=3 independent experiments, and bar graph represents the Mean ± S.E.M. of n=3 independent experiments.

Figure 3: Exogeneous IL-15 and IL-21 partially rescued granzyme B and activation-induced cell death, but not cytotoxicity of B cell-primed CTLs.

(A) Granzyme B expression in CTLs primed with B cells in the absence or presence of IL-15 and IL-21. (B and C) pMel-1 CTLs primed with mDCs, B cells or B cells in the presence of IL-15 and IL-21 were incubated for 6 hours with peptide-pulsed EL4 target cells, and assessed for (D) AICD of CD8+ CTLs, and (E) in vitro cytotoxicity. Representative histograms and bar graph depicting Mean ± S.E.M. from n=3 independent experiments.

Figure 4: Volcano plot of miRNA expression profiling.

The 18 miRNAs significantly differentially-expressed in DC- and B cell-primed CTLs are indicated in green; red line indicates p=0.05.

Figure 5: Heatmap of miRNAs differentially-expressed by DC- and B cell- primed CTLs.

After 3 days of in vitro priming by DCs or B cells, pMel-1 CTLs were isolated for miRNA expression profiling. Data shown is from n=3 independent miRNA profiling experiments.

Table 1: 18 miRNAs significantly differentially expressed in DC- and B cell primed CTLs, as determined by the paired t-test with significance level set at 0.05.

Figure 6: miR-23a inhibited both granzyme B and T-bet expression in CTLs in vitro.

DC-primed pMel-1 CTLs were retrovirally transduced with either an empty mock vector or a miR-23a overexpression vector. 3 days post-transduction, CD8+GFP+ CTL effector molecules expression was assessed by flow cytometry.

Figure 7: miR-23b, the paralog of miR-23a, did not affect Granzyme B and T-bet expression in CTLs.

pMel-1 CTLs were retrovirally-transduced with a Mock vector, the miR-23a overexpression vector or the miR-23b overexpression vector. 4 days post-transduction, Granzyme B and T-bet expression in transduced pMel-1 CTLs were assessed by flow cytometry. Numbers in histograms represent mean fluorescence intensity.

Figure 8: Validation of differential miR-23a expression following in vitro priming. miR-23a expression in CTLs primed under the respective conditions was verified with 3 additional batches of samples. Numbers and bar graph represents the Mean ± S.E.M. miR-23a expression relative to that of naïve CD8+ T cells.

Figure 9: miR-23a does not affect proliferation or AICD of activated CTLs.

pMel-1 CTLs were retrovirally-transduced with a Mock vector or the miR-23a overexpression vector. 4 to 6 days post-transduction, Mock and miR-23a-overexpressing TCRβ+GFP+ pMel-1 CTLs were restimulated for (F) 48 hours to assess cell proliferation or (G) 72 hours to assess AICD.

Figure 10: Quantification of miR-23a expression levels in activated murine CTLs.

(A) miR-23a overexpression levels in CTLs 4 days post-transduction from n=6 independent samples. (B) Chemically-synthesized miR-23a was serially diluted to generate a standard curve for miR-23a copy number determination. Linear regression was conducted and the R2 coefficients are shown. (C) miR-23a copy number in Mock- transduced pMel-1 CTLs from (A).

3.2.2 Forced miR-23a expression compromises anti-tumor CTL effector responses in vivo

To investigate the impact of miR-23a on CTL anti-tumor efficacy in vivo, we made use of the poorly-immunogenic B16/F10 melanoma tumor model (Overwijk et al.,

2003). 0.6×10 6 Mock pMel-1 CTLs, miR-23a overexpressing pMel-1 CTLs, or PBS vehicle control were infused into B16/F10 tumor-bearing mice. As previously reported

(Kerkar et al., 2011; Klebanoff et al., 2011), pMel-1 CTLs expressing the Mock vector retarded tumor growth substantially. However, this protection was completely abrogated by the forced expression of miR-23a – mice receiving miR-23a-overexpressing CTLs exhibited accelerated tumor progression and higher tumor burdens, comparable to that of untreated (PBS) tumor-bearing mice (Figure 11, A and B). Although miR-23a did not affect CTL accumulation within the tumor (Figure 11, C and D), miR-23a significantly undermined the expression of several key effector molecules in pMel-1 TILs (Figure

12A), and in peripheral pMel-1 CTLs (Figure 12). Taken together, we have functionally validated that forced miR-23a expression antagonizes anti-tumor CTL effector responses in vivo.

Figure 11: Forced expression of miR-23a in tumor-specific CTLs impairs tumor retardation, but does not alter their intratumoral accumulation.

3 days after subcutaneous inoculation of 0.2×10 6 B16/F10 melanoma cells, C57BL/6 tumor-bearing mice received adoptive transfers of 0.6×10 6 Mock or miR-23a overexpressing Thy1.1+ pMel-1 CTLs. Tumor progression was monitored and effector responses of Thy1.1+ pMel-1 CTLs were analyzed. (A) Tumor volumes of untreated mice (PBS), and mice adoptively transferred with Mock or miR-23a over-expressing Thy1.1+ pMel-1 CTLs. Graph represents Mean ± S.E.M. with n=9 mice per group, pooled from two independent experiments. *p<0.05 and ***p<0.001 for Mock VS miR- 23a; #p<0.05 and ###p<0.001 for Mock VS PBS by two-way ANOVA and Bonferroni post-test. 11 days after T cell transfer, tumors were excised and Thy1.1+ pMel-1 CTL numbers was assessed. Data shown is from one representative experiment. (B) Tumor sizes from mice in (A). (C) Percentages and (D) densities of tumor-infiltrating Mock and miR-23a-overexpressing pMel-1 CTLs.

Figure 12: Forced expression of miR-23a in tumor-specific CTLs inhibits their anti-tumor effector responses in vivo.

11 days after T cell transfer, tumor-bearing mice were sacrificed, and Thy1.1+ pMel-1 CTL effector function was assessed by flow cytometry. Effector molecules expression in (A) tumor-infiltrating and (B) splenic and miR-23a-overexpressing pMel-1 CTLs. Data shown is pooled from two independent experiments, with n=9 mice per treatment group. Normalized MFI was calculated by dividing the MFI of each sample by the average MFI’s of the Mock group for each experiment.

3.2.3 Functional blockade of miR-23a in CTLs augments their anti- tumor function in vitro

Having identified miR-23a as a repressor of CTL cytotoxicity, we developed two strategies for blocking miR-23a function: treatment with an anti-miR-23a locked nucleic acid (LNA) (Obad et al., 2011) and retroviral transduction of a miR-23a decoy vector

(Ebert et al., 2007).

In the first approach, a 6-fluorescein (6-FAM) fluorescent label conjugated to the

5’-end of the LNA facilitates monitoring the transfection efficiency, enables CTLs that had taken up the LNA to be distinguished as an FAM+ population (Figure 13A), and provides LNA-treated FAM- CTLs as an internal control for the specificity of LNA- mediated miR-23a inhibition. In comparison to the FAM - pMel-1 CTLs, the expression of

Eomes, T-bet and granzyme B were augmented in the miR-23a-inhibited FAM + population (Figure 13, B and C). To ensure that these observed changes are specific to miR-23a, CTLs were also treated with saturating amounts (Figure 14A) of a scrambled antagomir (Figure 14B), or an antagomir against the unrelated miR-122 (Figure 14C).

However, neither antagomir was able to augment CTL effector molecule expression. In spite of this functional enhancement, chemically-modified oligonucleotides, such as

LNAs, cannot be permanently-retained in activated T cells in vivo (data not shown and

(Obad et al., 2011)). The short-lived effects of LNA treatment in T cells therefore makes it difficult to investigate the long-term anti-tumor efficacy of miR-23a-inhibited CTLs in vivo.

Aimed at achieving long-lasting miR-23a inhibition, we developed a second approach, by retrovirally transducing CTLs with a miR-23a decoy vector capable of sequestering endogenous miR-23a (Ebert et al., 2007). To simultaneously allow the selection of engineered cells, we constructed a bicistronic viral backbone, in which the expression of a selectable marker (iRFP (Filonov et al., 2011) or puromycin-resistance) and a GFP decoy/reporter are driven independently by the viral 5’-LTR and PGK promoters, respectively (Figure 15A). To maximize their independent expression, we inserted an insulator sequence (Bell et al., 1999) between the selectable marker and decoy cassettes. A vector omitting miR-23a target sites serves as a Mock control. In CTLs transduced with the miR-23a decoy, GFP expression was substantially quenched by 85%, indicating that endogenous miR-23a had been sequestered by our synthetic 3’UTR

(Figure 15B). Consistent with LNA-mediated miR-23a suppression, the miR-23a decoy augmented the expression of cytotoxic modulators and effectors in CTLs (Figure 16A), and significantly enhanced their in vitro cytotoxicity over a wide range of E:T ratios

(Figure 16B), reiterating our findings that miR-23a inhibition effectively augments CTL functional capacity on a per-cell basis.

Figure 13: LNA-mediated functional blockade of miR-23a enhances CTL effector responses ex vivo.

Purified naïve pMel-1 CTLs were activated with αCD3/αCD28 for 48h in vitro, with or without 50nM FAM-tagged anti-miR-23a LNA. (A) Gating strategy to identify CTLs that have taken up the anti-miR-23a LNA. (B) CTL effector molecules expression in untreated (0nM LNA), FAM- LNA-treated (50nM LNA; FAM-) and FAM+ LNA-treated (50nM LNA; FAM+) CTLs. Representative histograms of n=6 independent experiments. (C) MFI of CTL effector molecules in FAM- and FAM+ CTLs treated with 50nM anti- miR-23a LNA; n=7. P-values were determined by two-tailed paired t-test.

Figure 14: Inhibiting miRNAs unrelated to miR-23a do not enhance CTL effector function.

Purified naïve pMel-1 CTLs were activated with anti-CD3/anti-CD28 for 48h in vitro, in the presence or absence of 50 μg/ml PE-tagged scrambled antagomir, or the unrelated miR-122 antagomir. (A) Gating strategy to identify Antagomir+ CTLs. Note all CTLs were Antagomir+, indicating saturating amounts of antagomirs. Expression of effector molecules in CD8+ CTLs treated with (B) scrambled antagomir and (C) miR-122 antagomir were assessed by flow cytometry, and compared to their untreated counterparts. Data shown is one representative of three independent experiments.

Figure 15: The miR-23a decoy retroviral expression vector.

(A) Schematic of the Mock decoy and miR-23a decoy retroviral expression vectors. iRFP was the internal marker for monitoring transfection efficiency; puromycin resistance (Puro R) was the selection marker for enriching engineered cells; GFP was a reporter for miR-23a sequestration and decoy function. The iRFP and GFP-decoy expression cassettes were separated by an insulator (Bell et al., 1999) ( ). (B) Representative dot plots of iRFP and GFP expression in pMel-1 CTLs transduced with the miR-23a decoy, where iRFP was used as the internal marker (right). Quenching of GFP intensity by the miR-23a decoy in CD8+iRFP+ CTLs (left). Bar graph represents Mean ± S.E.M.; n=6.

Figure 16: Decoy-mediated functional blockade of miR-23a enhances CTL effector responses ex vivo.

(A) pMel-1 CTLs retrovirally transduced with a Mock decoy vector or the miR-23a decoy vector were assessed for CTL effector molecules expression in vitro. P-values were determined by two-tailed paired t-test. (B) In vitro cytotoxicity of sorted iRFP+ Mock and miR-23a decoy-expressing OT-1 CTLs. Representative data of n=3 independent experiments.

3.2.4 miR-23a blunts CTL effector responses by targeting Blimp-1

We next sought to investigate the molecular mechanism through which miR-23a modulates CTL effector function. Since glutamine metabolism is central for appropriate

T cell activation (Wang et al., 2011), and miR-23a has previously been shown to directly target glutaminase (GLS) in cancer cell lines (Gao et al., 2009), we examined the impact of miR-23a on GLS expression in pMel-1 CTLs. However, inhibiting miR-23a by means of the miR-23a decoy failed to up-regulate Gls mRNA levels (Figure 17), indicating that miR-23a was unlikely to target Gls in primary effector CTLs. Therefore, we went on to search for miR-23a targets using miRecords (Xiao et al., 2009). miR-23a was computationally-predicted to target Blimp-1 (encoded by Prdm1 gene) at a highly evolutionarily-conserved site (Figure 18A), and to target T-bet (Figure 18B) and Eomes

(Figure 18C) at weakly-conserved sites.

To assess direct binding of miR-23a to the 3’-UTR of these predicted targets, we constructed luciferase reporters containing the full-length 3’-UTR of the Blimp-1, T-bet or Eomes genes. Each of these luciferase reporters was then co-transfected into Jurkat T cells, together with either a Mock vector or the miR-23a overexpression vector.

Luciferase activity controlled by the Blimp-1 3’-UTR, but not the T-bet or Eomes 3’-

UTRs, was significantly suppressed by miR-23a (Figure 19A). When the Blimp-1 3’-

UTR site predicted to interact with the miR-23a seed region was mutated (Figure 19B), luciferase reporter activity was no longer controlled by miR-23a (Figure 19C). Consistent with this, the miR-23a decoy rescued Blimp-1 mRNA (Figure 19D) and protein (Figure

19E) expression in pMel-1 effector CTLs. Interestingly, although T-bet is not a direct

target of miR-23a, T-bet mRNA levels were modestly, albeit significantly increased in miR-23a-inhibited CTLs. Mirroring earlier findings that T-bet transcription is down- regulated in Prdm1-/- effector CTLs (Rutishauser et al., 2009), the observed augmentation in T-bet expression is likely a secondary effect of increased Blimp-1 abundance in miR-23a-inhibited CTLs. Taken together, by directly regulating Blimp-1 expression, silencing miR-23a in CTLs increases their anti-tumor responses and cytotoxic potency.

Figure 17: miR-23a does not repress the expression of glutaminase in effector CTLs. pMel-1 CTLs were retrovirally-transduced with either the Mock decoy, or the miR-23a decoy vector. 3 days post-transduction, transduced CTLs were assessed for GLS mRNA expression. Bar graph represents Mean ± S.E.M. from n=2 independent experiments.

Figure 18: Computationally-predicted miR-23a targets relevant to CTL effector function.

(A) Schematic representation of the putative miR-23a binding site within the prdm1 3´UTR that is conserved across species. Computationally-predicted miR-23a target sites in the 3’-UTR of murine (B) T-bet and (C) Eomes.

Figure 19: Blimp-1 is a direct target of miR-23a in CTLs.

(A) Luciferase assays, in which Jurkat T cells were co-transfected with reporter constructs containing full-length 3’ UTRs of the indicated genes, together with the Mock or miR-23a overexpression vector. Graph represents Mean ± S.E.M.; n=5. (B) Mutant sequence of predicted miR-23a binding site in Blimp-1 3’UTR, with mutated nucleotides shown in bold. (C) Luciferase reporter assay, in which Jurkat T cells were co-transfected with reporter constructs containing the mutant Blimp-1 3’ UTR, together with the Mock or miR-23a overexpression vector. Graph represents Mean ± S.E.M. of n=3 independent experiments. (D and E) iRFP+ Mock and miR-23a decoy-expressing pMel-1 CTLs were sorted for miR-23a target studies. (D) Relative mRNA expression of the predicted miR- 23a targets Blimp-1, Eomes and T-bet, as well as other CTL effector molecules. Graph represents Mean ± S.E.M.; n=3. (E) Blimp-1 protein expression, with relative band intensities normalized to β-actin. Left: representative Western blot. Right: Pooled data from n=3 independent experiments. P-values were determined by two-tailed paired t-test.

3.2.5 In effector CTLs, TCR activation and TGFβ signaling differentially regulate miR-23a expression

Next, we sought to elucidate how signals received during priming, and within the tumor microenvironment, may reprogram effector CTLs through the alteration of miR-

23a levels. We initially identified miR-23a to be differentially-regulated in effector CTLs induced by different antigen-presenting cells (APCs) (Figure 5), indicating that miR-23a expression in CTLs can be modulated by cell-extrinsic signals. Therefore, we explored multiple pathways known to be differentially-influenced by DCs versus B cells. We first explored differences arising from the T:APC interface: TCR signaling strength, co- receptor signals and Notch signaling. Although TCR activation effectively suppressed miR-23a expression in CTLs, increasing the avidity of TCR signaling by varying plate- bound anti-CD3 antibody concentrations from 10ng/ml to 10 m g/ml did not further alter miR-23a levels (Figure 20). This 10ng/ml anti-CD3 threshold indicates that while miR-

23a expression is highly-sensitive to TCR activation, alteration of TCR signaling strength is not involved in fine-tuning miR-23a abundance in CTLs. Co-stimulatory signals from

CD28 and CD40L, too, had no effect on miR-23a expression (Figure 21, A and B). We also investigated the involvement of PD-1, an inhibitory checkpoint molecule exploited by immune-subversive tumors (Blankenstein et al., 2012; Curiel et al., 2003; Gilboa,

2004; Zou, 2005). Blocking the PD-1 ligands expressed on DCs (Figure 21C) had no effect on miR-23a abundance (Figure 21D). Conversely, inhibiting miR-23a in CTLs did not alter their surface expression of PD-1 (Figure 21E).

Since Notch signaling is known to promote CTL anti-tumor responses (Cho et al.,

2009; Sugimoto et al., 2010), and Notch ligands are differentially expressed on the surface of DC and B cells (Ohishi et al., 2003; Yoon et al., 2009), we speculated that

Notch activation may repress miR-23a expression. However, constitutively activating

Notch in CTLs by forced expression of the Notch1 intracellular domain (NICD) (Figure

22A) failed to impact miR-23a expression in CTLs (Figure 22B). In reciprocal loss-of- function studies, inhibiting Notch signaling with a γ-secretase inhibitor (Figure 22C) similarly had no effect on miR-23a expression in CTLs (Figure 22D).

In addition to cell:cell interactions, soluble cytokines generated during CTL priming may also influence miR-23a abundance. Therefore, we activated purified naïve

CTLs in the presence of various DC-derived cytokines for 3 days, before assessing miR-

23a expression. Among the panel of cytokines tested, which included Type 1 cytokines

(IL-2, IL-12, IFNγ and TNFα), inflammasome-derived cytokines (IL-1β and IL-18) and

Type 1 interferon (IFNβ), none were able to consistently or appreciably regulate miR-23a expression in CTLs (Figure 23A).

Finally, we explored the hypothesis that cytokines usually found within the tumor microenvironment – IL-6, IL-10 and TGFβ (Gilboa, 2004) – may promote miR-23a expression in CTLs. Within this group, IL-6 and IL-10 failed to appreciably impact miR-

23a expression (Figure 23B). In contrast, TGFβ up-regulated miR-23a levels (Figure

24A) and inhibited prdm1 expression (Figure 24B) in a dose-dependent manner.

Importantly, the regulation of prdm1 by TGFβ closely mirrored that of miR-23a: 1 ng/ml

TGFβ altered neither miR-23a nor prdm1 levels; however, a saturating dose of 10 ng/ml

TGFβ significantly stimulated miR-23a expression, while concurrently inhibiting the miR-23a target, prdm1. This indicates that the suppression of CTL cytotoxicity by TGFβ is, in part, post-transcriptionally mediated by miR-23a and its consequent suppression of the master regulator Blimp-1.

To investigate whether tumors are capable of driving up miR-23a expression in

CTLs, we activated CTLs in the presence of tumor cell-conditioned media. We additionally blocked TGFβ function with an antibody to directly assess the contribution from TGFβ. miR-23a was significantly up-regulated in CTLs treated with tumor cell- conditioned media; however, this increase was dampened upon neutralizing TGFβ

(Figure 24C). These demonstrate that in the context of the tumor microenvironment,

TGFβ is a primary modulator of miR-23a expression in anti-tumor CTLs.

Figure 20: TCR activation, but not stimulation strength, suppresses miR-23a expression in CTLs.

miR-23a expression in purified naïve CTLs activated in vitro with the indicated concentrations of αCD3 (μg/ml) and 5 μg/ml αCD28 for 3 days . Data represents Mean ± S.E.M.; n=4. *p<0.05 by one-way ANOVA and Bonferroni post-test.

Figure 21: Co-receptor signaling through CD28, CD40L and PD-1 do not affect miR-23a expression in CTLs.

(A) Purified naïve CTLs were activated in vitro with 5 μg/ml αCD3 and the indicated concentrations of αCD28 (μg/ml). Statistical analysis was performed by two-way ANOVA with Bonferroni post-test. (B) Naïve CTLs were primed in vitro with mDCs, in the presence of blocking antibodies against CD40L on T cells (MR1 antibody). (C) mDCs express the PD-1 ligands, PD-L1 and PD-L2. (D) Naïve CTLs were primed in vitro with mDCs, in the presence of blocking antibodies against PD-1 ligands (α-PD-L1 and α-PD-L2) on mDCs. After 3 days of priming, CTLs were purified for miR-23a expression analysis by qPCR. (E) miR-23a inhibition does not affect PD-1 expression on CTLs. CD8+iRFP+GFP+ pMel-1 CTLs transduced with the Mock or miR-23a decoy vector were assessed for surface expression of PD-1. (A) represent Mean ± S.E.M. from n=3 independent experiments and data shown in (B and D) represent Mean ± S.E.M. from n=2 independent experiments. Statistical significance in (B and D) was determined by the two-tailed paired t-test.

Figure 22: Notch signaling does not regulate miR-23a expression in CTLs.

(A and B) Notch activation was enforced in B cell-primed pMel-1 CTLs by transduction with a retroviral vector encoding the Notch intracellular domain (NICD), enabling constitutive activation of the Notch pathway. 4 to 6 days post-transduction, TCRβ+GFP+ pMel-1 CTLs were sorted for analysis of (A) mRNA expression of the Notch target gene Hes1; and (B) miR-23a expression. (C and D) Notch signaling was inhibited in mDC- primed pMel-1 CTLs by treatment with a γ-secretase inhibitor (GSI). After 3 days of in vitro priming, TCRβ+ pMel-1 CTLs were isolated and assessed for (C) Hes1 mRNA expression and (D) miR-23a expression. Data shown in represent Mean ± S.E.M. from n=2 independent experiments. ND: not detectable. Statistical significance was determined by the two-tailed paired t-test.

Figure 23: Cytokines that do not alter miR-23a expression in CTLs.

Purified naïve CTLs were activated with αCD3 and αCD28, in the presence of the indicated (A) inflammatory cytokines and (B) tumor-associated cytokines. After 3 days, CTLs were assessed for miR-23a expression. Data shown represents Mean ± S.E.M. from n>2 independent experiments. Statistical analysis was performed by one-way ANOVA with Bonferroni post-test.

100

Figure 24: TGFβ promotes miR-23a expression in CTLs.

(A) Mature miR-23a and (B) prdm1 expression in purified pMel-1 CTLs activated in vitro with varying concentrations of TGFβ for 72h. *p<0.05 by one-way ANOVA and Bonferroni post-test. (C) miR-23a expression in CTLs activated in vitro with tumor cell- conditioned media (25% in total medium) and neutralizing α-TGFβ antibody. Data shown represent Mean ± S.E.M.; n=3. Red dashed line represents expression levels in activated CTLs (control groups) that were set as 1.0, from which relative expression in experimental groups were calculated.

101

3.2.6 TCR and TGFβ signaling converge on cMyc to differentially modulate miR-23a expression in effector CTLs

We next investigated the signal integration from TCR activation and TGFβ stimulation that controls miR-23a expression. cMyc is one such convergent node impacted by both TCR (Bandukwala et al., 2012; Wang et al., 2011) and TGFβ receptor signaling (Genestier et al., 1999; Pietenpol et al., 1990). We examined whether cMyc plays a critical role in regulating pri-miR-23a expression in effector CTLs. Indeed, within

24h of activation, naïve pMel-1 CTLs rapidly up-regulate cMyc expression (Figure 25A), which coincides with a 6.7-fold decrease in pri-miR-23a transcription (Figure 25B). By contrast, during CTL priming, TGFβ repressed cMyc mRNA expression (Figure 26A), while augmenting transcription of the cMyc antagonists Mad1 and Mad4 (Figure 26B).

This finding in primary T cells parallels an earlier report that cMyc transcriptionally represses the precursor of miR-23a (pri-miR-23a) in cancer cell lines (Gao et al., 2009).

To interrogate the causal relationship between cMyc activity and miR-23a expression, we activated naïve pMel-1 CTLs in vitro in the presence of 10058-F4, a specific inhibitor that blocks Myc-Max dimerization (Yin et al., 2003). In TCR-activated CTLs, cMyc inhibition increased mature miR-23a expression (Figure 27A), but only resulted in a partial rescue (~50%) of pri-miR-23a transcripts (Figure 25B). At least two explanations may account for this incomplete rescue: one possibility is that other inhibitory mechanisms, in parallel to cMyc, may be involved in suppressing pri-miR-23a transcription; alternatively, transcribed pri-miR-23a may have undergone active miRNA processing, which prevents their accumulation and detection. With Dicer-deleted CTLs,

102

in which miRNA biogenesis was largely blocked, pri-miR-23a transcripts were fully rescued to levels of their naïve counterparts (Figure 27B), supporting the latter possibility. Taken together, these suggested that cMyc is the major repressor of pri-miR-

23a transcription in primed CTLs. Our findings thus identified cMyc as a key signaling node that integrates signals transduced through the TCR and TGFβ receptor to consequently govern miR-23a expression levels in effector CTLs. Interestingly, we noted that even with strong TCR signals, exposure of CTLs to TGFβ could effectively override

TCR-induced cMyc activation (Figure 26) and up-regulate miR-23a (Figure 24A).

Therefore, despite converging on the same signaling node, a TGFβ-enriched tumor microenvironment, but not tumor-antigen elicited TCR signaling, is the dominant regulator of miR-23a expression in CTLs.

103

Figure 25: TCR activation-induced cMyc represses pri-miR-23a transcription in effector CTLs.

(A) cMyc protein induction in purified CTLs upon 24h of TCR activation. (B) Pri-miR- 23a expression in purified pMel-1 CTLs treated with or without the cMyc inhibitor, 10058-F4. Data in (B) represent Mean ± S.E.M.; n=3. Red dashed line represents expression levels in activated CTLs (control groups) that were set as 1.0, from which relative expression in experimental groups were calculated.

104

Figure 26: TGFβ suppresses cMyc activity in effector CTLs. mRNA expression of (A) cMyc and (B) regulators of cMyc activity in activated CTLs upon TGFβ treatment. Data shown represent Mean ± S.E.M.; n=3.

105

Figure 27: cMyc is the major repressor of pri-miR-23a transcription in primed CTLs.

(A) Mature miR-23a expression in purified pMel-1 CTLs treated with or without the cMyc inhibitor, 10058-F4. Data in represent Mean ± S.E.M.; n=5. (B) Pri-miR-23a expression in Dicer-deficient CTLs upon cMyc inhibition, expressed as Mean ± S.E.M.; data represents two independent experiments, each performed in triplicates. Red dashed lines represent expression levels in activated CTLs (control groups) that were set as 1.0, from which relative expression in experimental groups were calculated.

106

3.2.7 Neutralizing miR-23a in CTLs mitigates TGFβ-induced immunosuppression

The secretion of TGFβ by malignant tumor cells poses a key hurdle to effective

CTL anti-tumor responses (Penafuerte and Galipeau, 2008; Zhang et al., 2013a). Having unveiled a novel regulatory pathway between TGFβ, miR-23a and Blimp-1 expression in

CTLs, we hypothesized this may be another mechanism by which TGFβ facilitates immune evasion; and, by functionally neutralizing this TGFβ-induced accumulation of miR-23a, we can preserve CTL immunocompetence. To directly validate this hypothesis, we interrogated the resilience of CTLs to TGF b challenge upon LNA- and decoy- mediated miR-23a functional blockade. In CTLs, IFNγ and granzyme B are known to be suppressed by TGFβ (Thomas and Massague, 2005), and are also cytotoxic effectors regulated by the miR-23a target, Blimp-1 (Kallies et al., 2009; Rutishauser et al., 2009).

Therefore, we used these two parameters as functional readouts of Blimp-1 rescue and cytotoxicity. Consistent with an earlier report (Thomas and Massague, 2005), TGF b dramatically repressed the acquisition of granzyme B (Figure 28A) and IFNγ (Figure

28B) by naïve CTLs; however, treatment with the anti-miR-23a LNA partially rescued the expression of these CTL effector molecules, even at high concentrations of TGF b

(Figure 28, A and B). As ACT utilizes activated, rather than naïve T cells, for tumor immunotherapy, we also sought to evaluate the functional competence of effector CTLs expressing the miR-23a decoy upon TGF b exposure. In agreement with earlier reports

(Ahmadzadeh M, 2005; Thomas and Massague, 2005), TGFβ treatment blocked IFNγ production in both Mock- and miR-23a decoy-transduced CTLs (Figure 28C). Again,

107

even after a 48h TGFβ conditioning, miR-23a decoy-transduced CTLs still maintained

IFNγ production at a level similar to TGFβ-untreated Mock CTLs. This indicates that miR-23a elevation is a major post-transcriptional mechanism through which TGFβ blunts

CTL cytotoxicity; importantly, inhibition of miR-23a in effector CTLs relieves TGFβ- mediated functional suppression.

108

Figure 28: Neutralizing miR-23a in CTLs mitigates TGFβ-induced immunosuppression.

Purified naïve pMel-1 CTLs were activated with αCD3/αCD28 and the indicated concentrations of TGFβ for 48h in vitro, in the presence or absence of 50nM FAM- tagged anti-miR-23a LNA. The percentage of (A) granzyme B- and (B) IFN-expressing TCRβ+CD8+FAM- and TCRβ+CD8+FAM+ cells were assessed by flow cytometry. Data shown represents Mean ± S.E.M.; of n=8 . ***p<0.001 by two-way ANOVA and Bonferroni post-test. (C) Mock- and miR-23a decoy-transduced pMel-1 CTLs were cultured with IL-2 for the first 48 hours, then washed and treated with the indicated concentrations of TGFβ for the next 48 hours. CTLs were restimulated with αCD3/αCD28 (1 μg/ml each) for the final 24 hours, and the percentage of IFNγ- producing CD8+iRFP+GFP+ cells was assessed by flow cytometry. Data shown represents Mean ± S.E.M.; of n=4. *p<0.05 and **p<0.01 by two-way ANOVA and Bonferroni post-test.

109

3.2.8 miR-23a expression correlates inversely with anti-tumor potential of mouse and human tumor-infiltrating CD8+ T cells

To understand the pre-clinical relevance of miR-23a expression in CD8+ T cells within the tumor microenvironment, we isolated CD8+ TILs from wildtype B6 mice bearing various burdens of B16/F10 melanoma. We found a strong positive correlation between tumor burden and miR-23a expression in CD8+ TILs (Figure 29A).

To assess whether miR-23a-mediated CD8+ T cell suppression is associated to the clinical pathology of human cancers, we explored the relationship between miR-23a expression and CD8+ TIL cytotoxic potential in a cohort of advanced lung cancer patients. In these patients, pleural effusion CD8+ T cells are a good reflection of CD8+

TILs, as they have received similar conditioning in the tumor microenvironment (e.g. by

IL-10 and TGFβ (Sikora et al., 2004)), and are known to be functionally-reminiscent of

CD8+ TILs (Atanackovic et al., 2004; Prado-Garcia et al., 2005). As a control for basal gene expression outside the tumor and to normalize for inter-individual variations, we also isolated CD8+ T cells obtained from the peripheral blood (PBMCs) of each patient for paired-sample analysis. As compared to CD8+ T cells in the periphery, miR-23a in

CD8+ TILs was elevated by a mean of 5.45±1.70 folds (Figure 29B). This up-regulation of miR-23a corresponded with a down-regulation of Blimp-1 mRNA levels in CD8+

TILs (Figure 30A); additionally, we found an inverse correlation between miR-23a and

Blimp-1 mRNA expression (Figure 30B). These observations reiterate our findings that miR-23a directly targets Blimp-1 in CD8+ T cells (Figure 19).

110

We also analyzed IFNγ mRNA and granzyme B protein expression as more direct read-outs of anti-tumor potential. In these patient samples, IFNγ expression was significantly down-regulated in CD8+ TILs (Figure 31A), and correlated inversely with miR-23a levels (Figure 31B); moreover, granzyme B expression on a population level and on a per-cell basis (Figure 31, C and D) were sharply diminished in CD8+ TILs.

Additionally, human PBMCs treated with the anti-miR-23a LNA showed enhanced granzyme B expression (Figure 32), indicating that functional blockade of miR-23a can boost the cytotoxicity of human CD8+ T cells. Taken together, we demonstrated that miR-23a is a clinically-relevant and translatable target for the immunotherapy of human cancers.

111

Figure 29: miR-23a is up-regulated in mouse and human CD8+ TILs.

(A) miR-23a expression in tumor-infiltrating CD8+ T cells (TILs) isolated from C57BL/6 mice bearing B16/F10 tumors. miR-23a expression in CD8+ TILs was normalized to basal expression of splenic CD8+ T cells in each individual mouse. Linear regression was conducted and the R 2 coefficient is shown. (B) CD3+CD8+ T cells from the pleural aspirates (TIL) and peripheral blood (PBMC) of advanced lung cancer patients were assessed for miR-23a expression levels. Relative miR-23a expression in patient CD8+ TIL (n=23) is expressed as a fold-change compared to patient-matched CD3+CD8+ PBMC. miR-23a was normalized to the average dCt of U6 and RNY3 endogenous controls. P-value was determined by the two-tailed paired t-test.

112

Figure 30: miR-23a levels correlate inversely with Blimp-1 expression in human CD8+ T cells.

CD3+CD8+ T cells from the pleural aspirates (TIL) and peripheral blood (PBMC) of advanced lung cancer patients were assessed for miR-23a and prdm1 mRNA levels. (A) Relative miR-23a and prdm1 expression in patient CD8+ TIL (n=10) are expressed as a fold-change compared to patient-matched CD3+CD8+ PBMC. miR-23a was normalized to the average dCt of U6 and RNY3 endogenous controls, while prdm1 mRNA was normalized to the average dCt of 18S RNA and RPLP0 endogenous controls. (B) Scatterplot of miR-23a VS prdm1 mRNA expression in CD3+CD8+ T cells from lung cancer patients (n=11). miR-23a dCt from each PBMC or TIL sample is plotted against its corresponding prdm1 mRNA dCt. The Pearson correlation coefficient (R) and two- tailed p-value are shown.

113

Figure 31: miR-23a expression correlates inversely with anti-tumor effector functions of human CD8+ T cells.

CD3+CD8+ T cells from the pleural aspirates (TIL) and peripheral blood (PBMC) of advanced lung cancer patients were assessed for miR-23a and cytotoxic effector molecules. (A) Relative IFNγ mRNA expression in patient CD8+ TIL (n=10) is expressed as a fold-change compared to patient-matched CD3+CD8+ PBMC. miR-23a was normalized to the average dCt of U6 and RNY3 endogenous controls, while IFNγ mRNA was normalized to the average dCt of 18S RNA and RPLP0 endogenous controls. (B) Scatterplot of miR-23a VS IFNγ mRNA expression in CD3+CD8+ T cells from lung cancer patients (n=10). miR-23a dCt from each PBMC or TIL sample is plotted against its corresponding IFNγ mRNA dCt. The Pearson correlation coefficient (R) and two- tailed p-value are shown. (C) Representative granzyme B histograms of CD3+CD8+ PBMC and TIL. (D) Percentage of granzyme B+ and granzyme B MFI from CD3+CD8+ PBMC and TIL (n=23). P-values were determined by the two-tailed paired t-test.

114

Figure 32: miR-23a blockade boosts the effector function of human CD8+ T cells.

Granzyme B expression in untreated (0nM) or LNA-treated (50nM) activated human CD3+CD8+ T cells from healthy donor PBMC. Numbers indicate granzyme B MFI. Histogram is representative of n=3 independent experiments.

115

3.2.9 Adoptive transfer of miR-23a-inhibited CTLs robustly retard tumor progression

In view of the clinical-relevance of miR-23a, we went on to examine the efficacy of our miR-23a targeted therapeutic strategy for cancer intervention. As a novel gene therapy tool, our bicistronic, dual-reporter retroviral construct (Figure 15A) poses several advantages for CTL programming: (i) it can be readily incorporated into conventional

ACT, as ex vivo-expanded tumor-specific CTLs can be simultaneously transduced with the retrovirus; (ii) the selectable marker enables successfully-engineered and functionally-robust CTLs to be selected/enriched for reinfusion; (iii) GFP reporter activity allows the effectiveness of miR-23a inhibition to be conveniently monitored.

With this tool, we mimic human cancer therapy by utilizing two mouse models of established tumors: B16/F10 melanoma (Figure 33) and Lewis lung cancer overexpressing ovalbumin (LLC-OVA) (Figure 34). When tumor masses reached palpable growth, we sub-lethally irradiated the tumor-bearing mice, and injected either

0.2×10 6 miR-23a Decoy-expressing pMel-1 or OT-1 CTLs intratumorally. Compared to equal numbers of Mock cells, treatment with miR-23a-inhibited CTLs dramatically retarded tumor progression (Figure 33A and Figure 34A), and significantly reduced tumor burdens (Figure 33, B and C; Figure 34, B and C). Upon examining tumor pathology 10 days post-CTL transfer, we found that although CTL persistence within the tumor mass was unaffected (Figure 35), miR-23a-inhibited CTLs showed augmented expression of the transcription factors T-bet and Eomes (Figure 36, A and B), and the cytolytic molecules IFNγ and granzyme B (Figure 36, C and D). Additionally, inhibiting

116

granzyme B accelerated tumor progression in vivo, and completely abrogated the anti- tumor advantage conferred by the miR-23a Decoy (Figure 37). Therefore, augmented granzyme B expression afforded by miR-23a-inhibited CTLs was functionally essential for enhanced melanoma clearance. Taken together, these indicated that suppressing miR-

23a in CTLs enhances their cytotoxic potency within the tumor microenvironment, thereby attaining optimal tumor eradication.

117

Figure 33:Adoptive T cell transfer therapy with miR-23a-inhibited CTLs dramatically retarded the progression of established tumors in a mouse model of melanoma.

7 days after subcutaneous inoculation of 0.2×10 6 B16/F10 melanoma cells, C57/BL6 tumor-bearing mice were sub-lethally irradiated and left untreated (PBS), or treated with intratumoral injections of 0.2x10 6 sorted iRFP+GFP+ Mock or miR-23a decoy-expressing pMel-1 CTLs. (A) B16/F10 tumor progression after the initiation of CTL therapy. Data represent Mean ± S.E.M. from n=6 mice per group in one representative of three independent experiments. **p<0.01 and ***p<0.001 indicate Mock VS miR-23a Decoy; ##p<0.01 and ###p<0.001 indicate Mock VS PBS by two- way ANOVA and Bonferroni post-test. B16/F10 (B) tumor sizes and (C) weight 10 days after CTL therapy.

118

Figure 34: Adoptive T cell transfer therapy with miR-23a-inhibited CTLs dramatically retarded the progression of established tumors in a mouse model of subcutaneous Lewis lung cancer.

7 days after subcutaneous inoculation of 0.2x10 6 LLC-OVA cells, C57/BL6 tumor- bearing mice were sub-lethally irradiated and left untreated (PBS), or treated with intratumoral injections of 0.2x10 6 sorted iRFP+GFP+ Mock or miR-23a decoy- expressing OT-1 CTLs. (A) LLC-OVA tumor progression after the initiation of CTL therapy. Data represent Mean ± S.E.M. with n=10 mice per group. *p<0.05 and **p<0.01 indicate Mock VS miR-23a Decoy; #p<0.05 and ###p<0.001 indicate Mock VS PBS by two-way ANOVA and Bonferroni post-test. LLC-OVA (B) tumor sizes and (C) weight 10 days after CTL therapy.

119

Figure 35: Intratumoral persistence of miR-23a-inhibited CTLs was unaffected.

10 days after adoptive T cell transfer, mice were sacrificed and tumors excised. Absolute numbers of Mock and miR-23a decoy-expressing Thy1.1+ pMel-1 CTLs within the tumor masses were analyzed with counting beads by flow cytometry. Data represent Mean ± S.E.M. from n=6 mice per group in one representative of three independent experiments.

120

Figure 36: miR-23a inhibition enhanced the anti-tumor effector function of tumor-specific CTLs in vivo.

10 days after adoptive T cell transfer, mice were sacrificed and tumors excised. Intratumoral Thy1.1+ pMel-1 CTLs isolated from B16/F10 tumors were analyzed for expression of the CTL master regulators and effector molecules (A) T-bet, (B) Eomes, (C) IFNγ and (D) granzyme B. Data represent Mean ± S.E.M., (A) from n=6 mice per group in one representative of three independent experiments.

121

Figure 37: Inhibiting granzyme B in pMel-1 CTLs abrogates the therapeutic advantage conferred by the miR-23a Decoy.

On Day 0, mice bearing large, established B16/F10 melanoma (>1000 mm 3) received intratumoral injections of 0.2×10 6 Mock Decoy pMel-1 CTLs, 0.2×10 6 miR-23a Decoy pMel-1 CTLs or 0.2×10 6 miR-23a Decoy pMel-1 CTLs pretreated with the granzyme B inhibitor zAAD-CMK. An additional 10 ug zAAD-CMK, or DMSO vehicle control, was injected intratumorally into each mouse on Day 3. Tumor volumes were normalized to that of Day 0, prior to the initiation of CTL therapy. Data represent Mean ± S.E.M. from n=4 (miR-23a Decoy + zAAD-CMK) or n=5 (Mock Decoy and miR-23a Decoy) mice per group. #p<0.05 and 0.01 for miR-23a Decoy VS Mock Decoy; *p<0.05 and **p<0.01 for miR- 23a Decoy VS miR-23a Decoy + zAAD-CMK by two-way ANOVA and Bonferroni post-test.

122

3.3 Discussion

CTL-based immunotherapy is a promising means of achieving durable control over tumor progression. However, its widespread use has been limited by the cost and effort in generating large numbers of anti-tumor CTLs ex vivo and by the incompetence induced by the tumor microenvironment. In this study, we attempted to redirect the focus of CTL engineering from amplifying the quantity to improving the quality of individual

CTLs, which could help to overcome both limitations. We initially identified miR-23a as a hurdle to effector CTL responses by differential priming with B cells or mature DCs.

Clearly, this in vitro priming system was not designed to recapitulate the complexities of

CTL responses within the tumor microenvironment; however, further investigation into factors that control miR-23a expression in CTLs led to our discovery that miR-23a was in fact a target of the immune-subversive tumor microenvironment. That miR-23a-inhibition imparts functional resilience to CTLs, particularly when challenged with immunosuppressive conditions, is supported by two pieces of evidence. Firstly, we observed that the enhancement in cytotoxicity provided by CTL-specific miR-23a blockade was consistently more profound in vivo (Figure 33A and Figure 34A) than in vitro (Figure 16B). We speculated that in vivo, the susceptibility of wildtype CTLs to

TGFβ-induced suppression might have magnified the functional advantage of miR-23a- inhibited CTLs. Secondly, our in vitro TGFβ-challenge experiments directly illustrated that inhibiting miR-23a in CTLs could, at least partially, preserve their immunocompetence in spite of high TGFβ concentrations (Figure 28).

123

Our mechanistic studies on miR-23a regulation uncovered a novel mechanism of

TGFβ-induced immunosuppression on CTLs: the TGFβ-miR-23a-Blimp-1 axis. The immunosuppressive effects of TGF b on CTLs are well-established (Ahmadzadeh M,

2005; Thomas and Massague, 2005; Zhang et al., 2013a). While earlier mechanistic studies focused on Smad2/3 residing on the cis-elements within regulatory regions of the ifng and gzmb genes (Thomas and Massague, 2005), our data revealed that even following Smad activation, there still exists a pathway of rescue to preserve cytotoxicity

(Figure 38). By targeting miR-23a, our in vitro and in vivo data demonstrated the robustness of this preservation, and, we attribute the robustness of miR-23a-mediated suppression to the strength of its target, Blimp-1. Blimp-1 is an essential master regulator that turns on the cytotoxic transcriptional program in activated CTLs. Notably, Blimp-1- deficient CTLs fail to differentiate into cytotoxic effectors, owing to their impaired expression of multiple cardinal cytotoxic molecules, including granzyme B and IFNγ, as well as the transcription factor T-bet (Kallies et al., 2009; Rutishauser et al., 2009). We showed, for the first time, that TGFb can control Blimp-1 expression through a miRNA- mediated posttranscriptional mechanism; moreover, abrogating miR-23a ameliorates

TGF b -induced CTL suppression, by rescuing the Blimp-1 downstream targets, granzyme

B and IFNγ (Figure 28). Our findings identify miR-23a as a TGF b -responsive rheostat that fine-tunes Blimp-1 levels in activated CTLs, and highlight the TGFβ-miR-23a-

Blimp-1 axis as a key posttranscriptional determinant controlling CTL cytotoxicity in an immunosuppressive environment. Therefore, TGF b -mediated immunosuppression is supported by at least two pillars: direct Smad-mediated transcriptional repression on 124

effector molecules, and indirect miR-23a-mediated posttranscriptional controls on the

Blimp-1. Most importantly from the therapeutic point of view, taking down just one pillar by blocking miR-23a function is sufficient to maintain CTLs’ cytotoxicity machinery at an adequate level for tumor intervention.

Within the tumor microenvironment, TGF b is a key mediator of tumor immune evasion. Blocking TGFβ signaling in CTLs – by TGFβ neutralization or enforced expression of the dominant-negative TGFβRII – can reverse their immune-tolerant state to promote tumor regression in vivo (Terabe et al., 2009; Thomas and Massague, 2005;

Ueda et al., 2009; Zhang et al., 2013a), making TGFβ and molecules in TGFBR- mediated signaling pathway drugable targets for tumor therapy (Connolly et al., 2012).

However, current pre-clinical and clinical data indicates that, due to its profound impact on immunosuppression and a wide range of physiological functions, systemic administration of anti-TGFβ reagents can cause severe inflammatory damage and other adverse off-target pathologies (Lonning et al., 2011). By contrast, during the process of ex vivo-expansion, ACT provides a window of opportunity to program tumor-specific

CTLs with immunocompetence against TGF b suppression. Of note, this reprogramming is restricted specifically to CTLs prior to reinfusion. Our findings highlight miR-23a as a clinically-relevant target for this purpose, whose functional blockade presents two significant advantages for ACT: it not only augments the cytotoxic potency of tumor- specific CTLs, but also mitigates TGFβ-induced immunosuppression.

125

Figure 38: Model of CTL immune-modulation by targeting miR-23a.

(A) Under immune-activating conditions, TCR signaling up-regulates cMyc in CTLs. Transcriptional repression of pri-miR-23a by cMyc permits Blimp-1 accumulation, resulting in increased expression of its cytotoxic target genes, granzyme B and IFNγ. (B) In the tumor microenvironment, TGFβ suppresses CTL activity via two mechanisms – Smad-mediated transcriptional reprogramming and miRNA-mediated post-transcriptional control. In the former, TGFβ-induced Smads are recruited to the gzmb and ifng gene regulatory regions to repress the transcription of these cytotoxic mediators directly. In the latter, TGFβ antagonizes cMyc activity, thereby derepressing pri-miR-23a transcription. Elevated miR-23a levels in CTLs down-regulates Blimp-1, and consequently its downstream cytotoxic effectors.

126

4. Extending the translational utility of ACT through CD8+ CAR T cells

4.1 Introduction

Cancer immunotherapy leverages the intrinsic defense mechanisms of host immunity not only to eliminate aberrant cells, but also to enforce long-term immune- surveillance. Adoptive T cell transfer (ACT) therapy seeks to accomplish this through the ex vivo expansion of autologous tumor-reactive immune cells, followed by reinfusion into the patient (Rosenberg et al., 2008). Due to their unique abilities to survey for and recognize aberrant target cells, execute cytotoxicity and generate long-lasting immunological memory, CD8+ cytotoxic T lymphocytes (CTLs) are functionally poised as vehicles for ACT cancer immunotherapy. Recent advancements in the field of T cell engineering, such as the introduction of chimeric antigen receptors (CARs) (Park et al.,

2011), have enabled polyclonal T cell specificity to be redirected to tumor antigens, further bolstering the utility of CTLs as biologics for ACT.

In spite of this renewed enthusiasm, ACT using CAR T cells has met with variable clinical success and considerable scientific controversy. Albeit effective for hematological malignancies (Maus et al., 2014), CAR therapy has at best achieved partial efficacy in the treatment of solid tumors (Kershaw et al., 2006; Lamers et al., 2006;

Morgan et al., 2006; Till et al., 2008). This is largely attributed to the susceptibility of

CAR T cells to the profoundly immune-suppressive tumor microenvironment (Moon et al., 2014). Moreover, due to the heterogeneous nature of tumors and the possibility for immune-editing, the single-antigen specificity of CARs raises concerns about their

127

application to tumors with heterogeneous expression and selective loss of the CAR- targeted antigen. Extending the clinical utility of CARs for ACT therefore warrants improved approaches that can instill functional resilience in CAR T cells, and calls for a mechanistic understanding of the long-term impact that CAR T cell therapy has on heterogeneous tumors.

To these ends, we performed pre-clinical studies utilizing a CAR that specifically recognizes EGFRvIII. EGFRvIII, a tumor-specific mutation of the epidermal growth factor receptor, is expressed in glioblastoma (GBM) and other common neoplasms

(Heimberger et al., 2005; Wikstrand et al., 1995; Wong et al., 1992), but not on normal tissues. The EGFRvIII mutant arises from an in-frame deletion of 801 basepairs encoding the extracellular ligand-binding domain of wildtype EGFR, bringing together amino acids

5 and 274 with the insertion of a glycine at the fusion junction (Sugawa et al., 1990;

Wong et al., 1992; Yamazaki et al., 1990). This generates the novel extracellular PepvIII epitope (LEEKKGNYVVTDH), an antigenic determinant unique to EGFRvIII

(Humphrey et al., 1990). EGFRvIII expression is associated with highly malignant GBM due to its tumorigenic properties: it acts as a constitutively active tyrosine kinase that confers tumorigenicity (Al-Nedawi et al., 2008; Inda et al., 2010), invasiveness

(Boockvar et al., 2003), and therapeutic resistance to tumor cells (Lammering et al.,

2004; Montgomery et al., 2000). The pre-clinical efficacy of EGFRvIII-CAR therapy for

GBM has been demonstrated in earlier studies: Human T cells transduced with

EGFRvIII-CAR not only exhibit antigen-specific cytotoxicity ex vivo (Morgan et al.,

128

2012), but also prolong the survival of immune-deficient mice bearing established human

GBM xenografts (Choi et al., 2014).

While encouraging, these initial studies present two major caveats. First, the cytotoxicity assays (Morgan et al., 2012) utilized peripheral blood T cells that were extricated from the GBM microenvironment, and were performed under culture conditions conducive for T cell activation. Therefore, these results did not take into account constraints imposed by the profoundly immune-suppressive GBM microenvironment in vivo (Jackson et al., 2011). Indeed, CAR therapy was not completely curative (Choi et al., 2014), and exposure to the tumor microenvironment impaired the effector function of tumor-infiltrating CAR T cells (Moon et al., 2014).

These observations indicate that despite having optimally-engineered T cell-activating moieties, CAR T cells still remain vulnerable to tumor-induced functional suppression in vivo. Second, examining CAR therapy in the context of an immune-deficient host (Choi et al., 2014) undermines the physiological and clinical relevance of the findings, as it discounts the involvement and complexities presented by an intact host immune system.

Subsequent studies in immune-competent hosts revealed that CAR therapy can, in fact, promote endogenous anti-tumor immune responses to mediate the clearance of antigen- heterogeneous tumors (Barber et al., 2009; Barber et al., 2008; Sampson et al., 2014;

Spear et al., 2013a). Specifically, EGFRvIII CAR-treated mice are resistant to subsequent re-challenge with PepvIII- GBM (Sampson et al., 2014), suggesting that EGFRvIII CARs can mobilize endogenous host immunity against additional tumor antigens (Sampson et al., 2014). Moreover, it was demonstrated that an intact endogenous T cell compartment

129

is essential for the optimal elimination of tumor cells and the development of anti-tumor immunological memory, indicating substantial crosstalk and synergism between CAR T cells and endogenous host immune cells (Spear et al., 2013b). However, the mechanism, immune cell subsets mobilized and involvement of endogenous immunological memory underlying this broad-spectrum, long-lasting anti-tumor efficacy delivered by CAR therapy remains elusive.

In this chapter, we seek to (i) investigate whether targeting miR-23a in EGFRvIII-

CAR T cells can enhance their immune-competence for GBM therapy, and (ii) interrogate the influence of EGFRvIII-CAR therapy on the functional reinvigoration of endogenous anti-GBM T cells.

4.2. Results

4.2.1 Targeting miR-23a in EGFRvIII-CAR T cells confers resistance to TGFβ-induced suppression

As described in Chapter 3, miR-23a inhibition increases cytotoxicity when T cells are stimulated through the T cell receptor. However, it is not known if the cytotoxicity of a 3 rd generation CAR ‒ whose triggering is simultaneously reinforced by both CD28 and

4-1BB co-stimulatory signals ‒ can be further enhanced with this approach. We addressed this issue by generating pMel-1 miR-23af/f ER-Cre mice, such that miR-23a deletion can be induced upon 4-hydroxytamoxifen (4-OHT) treatment in vitro, without perturbing neighboring miR-27a and miR-24 in this miRNA cluster (Figure 39). miR-

23a-deficient (miR-23aKO) and wildtype (WT) pMel-1 CD8+ T cells were then retrovirally transduced with the EGFRvIII-specific 3 rd generation CAR (based on the

130

human monoclonal antibody 139 scFv; Figure 40A) (Choi et al., 2014; Morgan et al.,

2012; Sampson et al., 2014), allowing their retargeting to the mutant PepvIII epitope of

EGFRvIII. Using a multimer reagent of biotinylated-PepvIII presented by streptavidin-

PE, surface expression of EGFRvIII-CAR was detected on transduced miR-23aKO and

WT pMel-1 CD8+ T cells (henceforth referred to as vIII-CAR 23aKO and vIII-CAR WT , respectively) (Figure 40B).

To investigate whether miR-23a abrogation in EGFRvIII-CAR T cells enhances their cytotoxicity and confers resistance to TGFβ, vIII-CAR 23aKO and vIII-CAR WT were restimulated with PepvIII peptide, together with various doses of recombinant TGFβ.

Their cytotoxic effector response, as reflected by IFNγ production, was then determined.

In the absence of TGFβ, stimulation through the CAR induced comparable amounts of

IFNγ in vIII-CAR 23aKO and vIII-CAR WT (Figure 41). However, in the presence of TGFβ,

IFNγ production by vIII-CAR WT was significantly attenuated (Figure 41). In contrast, miR-23a-deletion enabled EGFRvIII-CAR T cells to maintain high IFNγ expression in the face of TGFβ challenge: even with 10 ng/ml TGFβ ‒ the highest concentration tested

‒ vIII-CAR 23aKO still produced IFNγ at levels comparable to unsuppressed vIII-CAR WT

(Figure 41).

Our findings indicate that in the absence of immune-suppressive pressures, targeting miR-23a does not confer further enhancements to EGFRvIII-CAR T cell effector potency. This differs from our observations in conventional CD8+ T cells

(Chapter 3.2.3), and may be due to the saturation of T cell-activating signals provided by high-affinity antigen-binding, as well as CD28- and 4-1BB-mediated co-stimulation,

131

which cannot be further augmented by miR-23a deletion. Nevertheless, EGFRvIII-CAR

T cells remain vulnerable to TGFβ-induced suppression, indicating that the simultaneous co-stimulatory signals delivered by CD28 and 4-1BB in the 3 rd generation CAR are still insufficient to prevent TGFβ-induced functional suppression. Our findings are consistent with earlier reports utilizing the 2 nd generation CAR, which demonstrated that immune- barriers imposed by the tumor microenvironment can render those CAR T cells dysfunctional in vivo (Moon et al., 2014; Zhang et al., 2013a). Importantly, abrogating miR-23a in EGFRvIII-CAR T cells conferred resistance to TGFβ, indicating that this functional rescue afforded by miR-23a abrogation operates independently of CAR- mediated signaling. Targeting miR-23a in CAR T cells therefore represents a unique means of preserving CAR T cell immune-competence, particularly in the TGFβ-enriched microenvironments as in GBM (Bodmer et al., 1989).

132

Figure 39: miR-23a deletion efficiency in pMel-1 ER-Cre+ miR-23af/f CD8+ T cells.

Total lymph node cells from pMel-1 ER-Cre+ miR-23af/f mice were activated with 5 μM hgp10025-33 in vitro, in the presence of ethanol (EtOH) vehicle control (wildtype; WT) or 200 nM 4-OHT (miR-23a-deleted; KO) for 48 hours. CD8+ T cells were then purified and assessed for the expression of miR-23a-27a-24 cluster members by qPCR. Expression of each miRNA species in KO T cells relative to that of WT T cells are shown. Graph represents Mean ± S.E.M from n=2 mice.

133

Figure 40: Transduction of pMel-1 CD8+ T cells with the 3rd generation EGFRvIII-CAR retroviral vector.

(A) Schematic of the EGFRvIII-CAR retroviral vector. The 3 rd generation EGFRvIII- CAR couples GBM TSA recognition to T cell activation. The extracellular antigen recognition domain of EGFRvIII-CAR is a single chain variable fragment (scFv) based on the human monoclonal antibody 139, which is specific for the mutant PepvIII epitope of GBM-specific EGFRvIII. A transmembrane domain couples the scFv to intracellular T cell-activating moieties comprised of the mouse CD28, 4-1BB and CD3ζ signaling domains. (B) Expression of EGFRvIII-CAR in pMel-1 CD8+ T cells. Activated WT (ethanol-treated) and KO (4-OHT-treated) pMel-1 lymphocytes were retrovirally transduced with the EGFRvIII-CAR vector in (A). 2 days post-transduction, CD8+ T cells were analyzed by flow cytometry for surface expression of the EGFRvIII-CAR using a PepvIII-PE multimer. Untransduced CD8+ cells served as a background control to identify CAR+ populations.

134

Figure 41: miR-23a deletion in EGFRvIII-CAR T cells confers resistance to TGFβ-induced functional suppression.

WT and KO pMel-1 CD8+ T cells were transduced with EGFRvIII-CAR. 2 days post- transduction, cells were restimulated on plate-bound PepvIII peptide for 24 hours with the indicated concentrations of TGFb, before intracellular IFNγ expression in CD8+ EGFRVIII-CAR+ cells was assessed by flow cytometry. Graph represents Mean ± S.E.M from n=5 independent experiments. n.s denotes not statistically signficant, *p<0.05 and **p<0.01 by two-way ANOVA and Bonferroni post-test.

135

4.2.2 EGFRvIII-CAR therapy drives the mobilization and convergence of endogenous intratumoral T cell repertoires

Broad-spectrum immunological memory against diverse tumor antigens is essential for effective tumor eradication and long-term immune-surveillance. Previous data suggested that EGFRvIII-CAR therapy can mobilize endogenous immunity against

PepvIII-negative re-transplanted tumors. However, it is unknown whether this immune- protection is mediated by the mobilization of endogenous tumor-reactive T cells by epitope spreading; and if so, whether endogenous T cell clones mobilized in different individuals bear specificity for the same tumor antigens.

We therefore sought to determine the impact of EGFRvIII-CAR therapy on the expansion and diversity of endogenous GBM-infiltrating T cell clones. To this end, we utilized syngenic murine KR158B cells stably expressing luciferase and EGFRvIII

(KLucvIII; Figure 42) in an immune-competent model of murine GBM. Briefly,

KLucvIII implanted intracerebrally (IC) into wildtype B6 mice were allowed to establish for 14 days, followed by IC treatment with either saline, vIII-CAR 23aKO pMel-1, or vIII-

CAR WT pMel-1 CD8+ T cells. Of note, GBM-bearing mice did not receive additional lymphodepleting regimens or supportive cytokine therapy. 10 days post-CAR transfer, bioluminescence imaging (BLI) of KLucvIII revealed that compared to saline-treated controls, IC-CAR treatment had not yet elicited significant differences in the tumor burdens (Figure 43). Immediately following BLI, matched GBM and spleen samples from all mice were harvested and subjected to deep sequencing of TCRβ regions. By

136

identifying CAR T cells by the transgenic pMel-1 TCRβ gene sequence, we were able to exclude CAR T cells from the dataset and restrict our analysis to the endogenous T cells.

We first examined the distribution of endogenous T cell clones in the brain and the periphery. We observed that the GBM-bearing state alone is associated with the loss of immune-privilege in the CNS (Engelhardt, 2008): even in the absence of CAR therapy, saline-treated brains contained substantial numbers of distinct T cell clones that preferentially accumulated and expanded in the brain (Figure 44; blue dots below the diagonal). This indicated that the blood-brain-barrier (BBB) was highly permeable to T cells, and that a two-way traffic of T cells between the CNS and periphery was possible.

We therefore questioned how effectively IC-delivered CAR T cells could be retained in the tumor-bearing brain. To this end, we examined the in vivo distribution of IC- delivered EGFRvIII-CAR T cells in these mice. The vIII-CAR 23aKO treatment group was excluded from these analyses, as vIII-CAR 23aKO cells could not be detected in 2 mice and were only detected at low frequencies in the other 3 mice (Figure 44). We therefore focused our attention on the vIII-CAR WT treatment group, in which the adoptively transferred cells were present in sufficient numbers for analysis. Significant frequencies of IC-delivered vIII-CAR WT were detected among splenocytes, indicating they had escaped the BBB into the periphery (Figure 44; red dots). Nevertheless, vIII-CAR WT selectively accumulated in the brains of all mice in this treatment group, and represented the most dominant clonotype that occurred at the highest frequency (Figure 44; red dots).

The preferential accumulation of vIII-CAR WT in the GBM-bearing brain was possibly due to the local presence of the vIII-CAR-specific antigen EGFRvIII, supporting the

137

prospect that vIII-CAR WT may persist and expand in vivo for their effector responses to be triggered in response to EGFRvIII+ GBM.

Next, we examined the occurrence of highly-expanded clones (HECs), which we define as clones with frequencies greater than 0.5%. As expected, no spleen-enriched

HECs were detected in any of the mice (Figure 44; clones above horizontal green dashed lines). A low number of HECs was enriched in the brains of saline-treated mice (Figure

44; Saline; clones to the right of vertical green dashed lines); however, the number of brain-enriched HECs was increased following vIII-CAR WT delivery (Figure 44; vIII-

CAR WT ; clones to the right of vertical green dashed lines). This indicated that CAR therapy may promote the expansion of endogenous tumor-infiltrating T cell clones.

To measure the amount of similarity in the endogenous TCR repertoires between different mice within the same treatment group, we calculated the Bhattacharyya coefficients for TCRβ NT and CDR3 AA sequences of each group. As expected, splenic

TCR repertoires had low Bhattacharyya coefficients, reflecting the low inter-individual similarity of T cell clones in the periphery of each mouse (Figure 45, A and B). In contrast, as compared to the spleen, all three treatment groups had intratumoral TCR repertoires with significantly higher Bhattacharyya coefficients (Figure 45, A and B).

This indicated that tumor antigens drove the preferential selection and expansion of a subset of overlapping endogenous T cell clones, resulting in the convergence and increased similarity of inter-individual intratumoral TCR repertoires. Interestingly, we found that the vIII-CAR WT - treated group had the highest intratumoral Bhattacharyya coefficients (Figure 45, A and B), suggesting that vIII-CAR WT CAR therapy may

138

facilitate the mobilization of additional shared tumor-reactive T cell clones that may otherwise remain dormant in saline-treated mice.

We next performed unsupervised hierarchical cluster analysis to evaluate the intratumoral repertoire similarity between individual mice across all treatment groups.

Strikingly, the intratumoral repertoires of vIII-CAR WT - treated mice formed a distinct cluster, showing the greatest convergence (Figure 46). This further reiterates that vIII-

CAR WT therapy synchronously mobilizes a pool of endogenous tumor-reactive T cell clones that are shared in common by individual mice.

139

Figure 42: EGFRvIII expression on the KLucvIII murine GBM cell line.

The KLucvIII cell line and its KLuc parental line were analyzed for surface expression of mutant EGFRvIII by flow cytometry.

140

141

Figure 43: Comparable tumor burdens at point of tissue sample collection.

2×10 4 KLucvIII cells were implanted into C57BL/6 mice IC on Day -14. On day 0, tumor-bearing mice received IC injections of either saline, 2×10 6 EGFRvIIICAR- transduced WT pMel-1 cells (vIII-CAR WT ), or EGFRvIIICAR-transduced miR-23aKO pMel-1 cells (vIII-CAR 23aKO ). On Day 10, KLucvIII burdens were imaged by in vivo bioluminescence. (A) Bioluminescent images and (B) graph of luciferin signal expressed as geometric mean ± 95% CI. n.s.: not statistically significant by one-way ANOVA and Bonferroni post-test; ROI: region of interest; n=5 mice per treatment group.

142

143

Figure 44: Distribution and clonal frequencies of IC-delivered CAR T cells and endogenous T cells.

Scatterplot shows the TCRβ NT clonotype distribution between the brain (x-axis) and spleen (y-axis) in saline-treated, vIII-CAR WT -treated and vIII-CAR KO -treated GBM- bearing mice. Each scatterplot represents one mouse. Red dots represent NT sequences encoding the pMel-1 TCRβ gene; blue dots represent NT sequences encoding all other TCRβ genes. Dots falling within the coordinate represent shared clones detected in both brain and spleen; dots on the axes represent unique clones detected only in the respective tissue compartment. Dots along the diagonal represent clones that occur at similar frequencies in both tissue compartments; dots above and below the diagonal represent clones that are enriched in the spleen and brain, respectively. Green dashed lines demarcate a clonal frequency of 0.5% in the spleen (horizontal) and brain (vertical).

144

Figure 45: Tumor antigens drive the local accumulation of shared tumor- reactive T cell clones.

TCR repertoires from the brain tissue and splenocytes of each mouse was compared to that of every other mouse within the same treatment group. The Bhattacharyya coefficients for (A) TCRβ NT and (B) CDR3 AA sequences of each treatment group was then calculated. Br.: brain tissue; Sp.: splenocytes. P-values were calculated by the two-tailed Wilcoxon matched pair test, with n=5 mice per treatment group.

145

Figure 46: vIII-CAR WT therapy drives the convergence of inter-individual intratumoral TCR repertoires.

Unsupervised hierarchical cluster analysis of intratumoral TCR repertoires of individual mice across all treatment groups. Red box highlights the distinct clustering of intratumoral TCR repertoires from vIII-CAR WT -treated mice.

146

4.3 Discussion

GBM, the most common primary malignant brain tumor, is uniformly and rapidly lethal. With standard therapy, the median time to recurrence stands at a mere 6.9 months

(Stupp et al., 2005). Patient relapse is largely ascribed to the invasive and disseminate nature of GBM, underscoring the importance of generating broad spectrum, post-surgical immune-surveillance to eradicate residual tumor cells and to prevent recurrence.

Immunotherapy with EGFRvIII-CAR, which targets the tumor-specific PepvIII epitope, has promised a means to this end. In spite of its immense therapeutic potential, the extensive clinical use of EGFRvIII-CAR therapy has been hampered, potentially by the susceptibility of CAR T cells to immune-suppression in vivo, as well as a general lack of understanding on how these infused CAR T cells impact the immune status of patients in the long-run. In the hopes of broadening the practical applications and defining the mechanistic consequences of CAR therapy, we performed preliminary studies employing the use of genetically-engineered tumor-specific CD8+ T cells

In Chapter 4.2.1, we provided the proof-of-concept that miR-23a abrogation in

CD8+ EGFRvIII-CAR T cells afforded them with functional resilience to TGFβ- mediated immune-suppression. Mechanistically, we have identified TCR activation and

TGFβ as opposing signals that differentially regulate miR-23a in CD8+ T cells (described in Chapter 3.2.6). Integrating these findings leads to our conclusion that despite possessing optimal T cell stimulatory capacities, strong T cell activation signals delivered through the CAR still remains inept at suppressing TGFβ-induced miR-23a expression.

GBM microenvironmental TGFβ therefore supersedes CAR-mediated T cell activation to

147

up-regulate miR-23a, resulting in partial therapeutic efficacy. Intercepting TGFβ-induced

CAR T cell suppression through the inhibition of miR-23a therefore represents a compelling means for enhancing CAR-based ACT.

An intriguing observation that arose from our in vivo studies (Chapter 4.2.2) was the extreme permeability of the BBB induced by the GBM-bearing state, which allowed an active two-way traffic of T cells between the periphery and the CNS. Although the presence of brain-infiltrating T cells has been reported to be a common phenomenon among GBM patients, they are generally thought to occur at low numbers (Han et al.,

2014; Kim et al., 2012). Surprisingly, we detected the presence of substantial numbers of unique endogenous T cell clones that preferentially expanded within GBM brain tissues

(Figure 44; blue dots under the diagonal in Saline treatment group), demonstrating that the BBB is more permeable to peripheral T cell infiltration than previously imagined. In contrast, the tumor-bearing BBB has previously been reported to be poorly-permeable to antibodies and chemotherapeutic drugs (Lampson, 2011; Lockman et al., 2010). That the

BBB is highly-permeable to T cells suggests that the trafficking properties of T cells poise them as ideal vehicles for the immunotherapy of CNS malignancies.

In addition, our findings that the leaky GBM BBB permits CAR T cells to escape the CNS into the periphery ‒ despite brain-targeted IC delivery ‒ has important implications for the rational design of GBM targets for immunotherapy. Besides the

EGFRvIII-CAR targeting the GBM-specific PepvIII epitope, other CARs that recognize

GBM-associated antigens have also been developed. These include CARs specific for anti-IL-13Rα (Kong et al., 2012), human epidermal growth factor receptor 2 (HER2)

148

(Ahmed et al., 2010) and erythropoietin-producing hepatocellular carcinoma A2 (EphA2)

(Chow et al., 2013), all of which have proven efficacious in preclinical models of murine

GBM and are currently undergoing further clinical development. Nevertheless, these targeted antigens are also expressed by normal tissues in the periphery. Coupled to our observations that even IC-delivery is unable to anatomically restrict CARs to the brain, the escape of these TAA-targeted CARs to the periphery to cause destruction of normal tissues becomes a real threat. We therefore caution against the use of TAA-targeted

CARs, and urge the pursuit of TSA’s to enhance the safety profile of CAR-based ACT.

Pre-clinical studies in mouse models of GBM have demonstrated that EGFRvIII-

CAR therapy can cause the regression of orthotopic PepvIII+ GBM, and more importantly, can prevent subsequent relapse from PepvIII- tumors (Sampson et al., 2014).

While these findings greatly extend the potential utility of CAR T cells for the treatment of antigen-heterogeneous tumors, the immunological mechanisms underlying this broad- spectrum protection remain elusive. In our studies, we seek to provide a plausible and compelling explanation through clonotypic analysis of endogenous T cells following

CAR therapy. Our data revealed that vIII-CAR WT treatment increased the overlap of inter-individual TIL repertoires (Figure 45), suggesting that CAR therapy may lower the threshold for the intratumoral mobilization of sub-dominant tumor-reactive T cell clones by epitope spreading. Based on our findings, we postulate that CAR therapy induces endogenous T cell epitope spread to additional tumor antigens; in this way, initial single- target therapy with CARs can subsequently kick-off a diverse endogenous T cell response

149

against multiple tumor antigens, thus broadening the spectrum of immune-recognition and surveillance critical for protection against heterogeneous tumors.

In addition, our finding that CAR therapy drives the convergence of inter- individual TIL repertoires may bear important implications for the design of future immunotherapies. This, however, requires further investigation to verify if TIL repertoire convergence also holds true in CAR-treated patients sharing the same MHC alleles. If it does, we may be able to identify “public” T cell clones – where T cells having the same

CDR3 AA sequences are elicited in different cancer patients (Venturi et al., 2008). These public TCRs are compelling therapeutic tools: their highly frequent occurrence in multiple individuals indicates that they are unlikely to be negatively selected during development, and are likely to be specific for “non-self” antigens that are commonly expressed by tumor cells. This will spur the engineering of novel TCRs that are less self- reactive and more tumor-specific, thus enhancing the safety and efficacy of TCR gene therapy for cancer. Moreover, identifying the antigenic epitopes recognized by these public TCRs will help uncover novel tumor-specific antigens for therapeutic targeting.

More detailed analysis of the endogenous TCR β deep sequencing results are currently underway, with which we hope to better characterize how CAR therapy impacts the clonal expansion and diversity of endogenous T cells. Taken together, our studies in this Chapter have attempted to enhance the efficacy of CAR therapy by manipulating

CD8+ T cells, and sought to examine the extent of immunological protection conferred by CAR therapy.

150

5. Biological evaluation of subglutinol A as a novel immunosuppressive agent for the intervention of T cell- mediated inflammatory diseases

The contents in this chapter were originally published in ACS Medicinal

Chemistry Letters. Biological evaluation of subglutinol a as a novel immunosuppressive agent for inflammation intervention. Regina Lin, Hyoungsu Kim, Jiyong Hong, and Qi-

Chemical Society. (http://pubs.acs.org/doi/abs/10.1021/ml4004809)

5.1 Introduction

Subglutinols A and B (Figure 47) are immunosuppressive natural products isolated from Fusarium subglutinans, an endophytic fungus from the vine Tripterygium wilfordii (Lee et al., 1995; Strobel and Pliam, 1996). T. wilfordii has long been used as an anti-inflammatory in traditional Chinese herbal medicine (Ho and Lai, 2004; Tao and

Lipsky, 2000). In addition, formulated extracts made from T. wilfordii are Chinese FDA approved drugs (Lei Gong Teng tablets, CFDA approval #Z42021534) for rheumatoid arthritis, psoriasis, lupus associated autoimmune nephrotic syndrome, and autoimmune hepatitis. Subglutinols A and B were highly potent in the mixed-lymphocyte reaction

(MLR) and thymocyte proliferation (TP) assays (IC50 0.1 μM) (Lee et al., 1995; Strobel and Pliam, 1996). Owing to the lack of toxicity (Lee et al., 1995; Strobel and Pliam,

1996), 1 and 2 were expected to be promising new immunosuppressive drugs and have attracted a strong interest (Kikuchi et al., 2011; Kim et al., 2009; Kim et al., 2010; Lee et al., 2012). Subglutinol A (IC50 = 25 nM) was indeed more potent than CsA (IC50 = 89

151

nM) in the MLR assay (Kim et al., 2009; Kim et al., 2010), suggesting its potential for use as an immunosuppressive pharmacologic agent. Moreover, the total synthesis of subglutinols A and B can now be achieved by chemical synthesis from organic precursors

(Kim et al., 2009; Kim et al., 2010); this not only facilitates analogue development for further structure-function optimization, but also makes possible the controlled production of high-purity products for pharmacological applications. However, the efficacy, mode of action and tolerability of systemically administered subglutinols have yet to be characterized.

Encouraged by the promising preliminary data, we embarked on more comprehensive evaluation of in vitro and in vivo immunosuppressive activity of subglutinol A. Herein, we report the efficacy of subglutinol A in eliminating Th1 and

Th17 responses in vitro and in suppressing inflammation in vivo to demonstrate the potential of subglutinol A as a novel immunosuppressive agent for autoimmune diseases.

152

Figure 47: Structure of subglutinols A and B.

153

5.2 Results

5.2.1 Subglutinol A inhibits the expansion of activated T cells

As the commander controlling adaptive immune responses, T lymphocytes are tightly-restricted to a quiescent state under normal conditions. During an infection, foreign antigen-specific T cells are rapidly activated and cycle to exponentially increase their numbers, which is essential for the efficiency of pathogen clearance. Upon resolution of the infection, most of these activated T cells are eliminated by apoptosis to keep the immune response in check and to prevent bystander tissue damage (Vigano et al., 2012). During the onset of autoimmune diseases, however, self-antigens are recognized as non-self and elicit immune responses from auto-reactive T cells. Mirroring the immune responses towards foreign pathogens, massive clonal expansion and failure of contraction lead to the accumulation of pathogenic T cells responsible for mediating autoimmune diseases. In fact, defects in T cell apoptosis (e.g., resistance to Fas-induced cell death) are etiological causes of autoimmune lymphoproliferative syndrome (ALPS) and multiple sclerosis in humans (Comi et al., 2012; Fleisher and Oliveira, 2012).

Therefore, developing therapeutics capable of modulating the proliferation and cell death pathways of autoreactive T cells is a compelling strategy for the treatment of autoimmune disorders (Fife and Pauken, 2011; Pender, 2007).

To comprehensively evaluate the capacity of subglutinol A in suppressing T cells’ effector responses, we used primary mouse T cells from the pMel-1 and LLO118 T cell receptor (TCR) transgenic mice. pMel-1 mice have CD8+ T cells that are specific for the melanoma antigen glycoprotein 100 (hgp100 25-33 ), while LLO118 mice bear CD4+ T

154

cells recognizing the Listeria toxin listeriolysin (LLO 190-205 ) (Overwijk et al., 2003;

Weber et al., 2012). Naïve pMel-1 CD8+ T cells and LLO118 CD4+ T cells were activated with their respective peptide antigens (5 μM hgp100 25-33 or 10 μM LLO 190-205 ) for 48 h in vitro, in the presence of 100 nM of subglutinol A or cyclosporine A (CsA).

DMSO was included as the vehicle control. T cell proliferation was assessed by carboxyfluorescein diacetate succinimidyl ester (CFSE) dilution and the cell death was determined by Live/Dead staining. During the early phase of antigen-induced T cell activation, for both CD4+ and CD8+ T cells, subglutinol A reduced their survival ratio from 80% to ~10% (Figure 48). Among the remaining live T cells, cell division was completely halted. In a side-by-side comparison to CsA, subglutinol A has a slightly lower cytotoxicity for undivided T cells, but has a better efficacy in inhibiting T cell proliferation. Therefore, by intercepting these initial fundamental events of T cell activation, subglutinol A has the potential to prevent the accumulation of auto-reactive pathogenic T cells.

155

Figure 48: Subglutinol A blocks antigen-induced T cell proliferation and induces massive apoptosis.

CFSE-labeled naı̈ve pMel-1 CD8+ (top) and LLO CD4+ transgenic T cells (bottom) were activated in vitro, in the presence of 1 (100 nM), CsA (100 nM), or DMSO vehicle control. After 48 h, T cell proliferation and cell death were assessed by CFSE dilution and live/dead staining, respectively. Data shown is representative of two independent experiments.

156

5.2.2 Subglutinol A abrogates IL-2 production by activated CD4+ T cells

During an immune response, activated CD4+ T cells are the primary producers of

IL-2, a pleitrophic cytokine that influences multiple immune cell subsets. Not only is IL-

2 a potent growth factor essential for T cell proliferation and survival, but it also drives the effector differentiation of Th1 cells, cytotoxic CD8+ T cells and NK cells that perpetuate tissue destruction (Boyman and Sprent, 2012; Lan et al., 2008). Therefore, we sought to investigate the influence of subglutinol A on CD4+ T cell-derived IL-2 production. Naïve LLO118 CD4+ T cells were activated in vitro under a Th0 (non- polarizing) condition for the first 4 days, during which they acquired IL-2-competence. In the 48 hours that followed, these activated IL-2-competent T cells were then treated with various doses of subglutinol A, CsA, or the vehicle control DMSO. Cytokine production was evaluated by intracellular IL-2 staining and flow cytometry. Among DMSO-treated

CD4+ T cells, over 30% of cells were capable of producing IL-2 (Figure 49A). Though subglutinol A was less potent at inhibiting IL-2 production than CsA (Figure 49, A and

B), subglutinol A still effectively inhibited IL-2 expression by effector CD4+ T cells: IL-

2 production fell around 50% upon treatment with 1 nM of subglutinol A and was completely abrogated at a higher dose of 100 nM (Figure 49, A and B).

157

158

Figure 49: Subglutinol A abrogates antigen-induced IL-2 production by activated CD4+ T cells.

LLO CD4+ T cells were activated by the cognate antigen in vitro for the first 4 days in the absence of drugs, followed by 2 additional days in the presence of the indicated concentrations of subglutinol A, CsA, or DMSO vehicle control (0 nM drugs), and assessed for IL-2 expression by intracellular staining and flow cytometry. (A) Numbers in dot-plots represent the percentages of IL-2-producers among live CD4+ T cells. Data shown is one representative of three independent experiments. (B) The percentage of IL- 2 inhibition was calculated. Data shown is Mean ± S.E.M of three independent experiments.

159

5.2.3 Subglutinol A abolishes pro-inflammatory cytokine production by effector Th1 and Th17 cells

As key drivers and perpetuators of inflammation, self-antigen elicited IFNγ and

IL-17 production by pathogenic Th1 and Th17 CD4+ T cells is a hallmark of autoimmune disorders, including Type 1 diabetes (Leung et al., 2010; Moudgil and

Choubey, 2011), rheumatoid arthritis (Boissier et al., 2012), multiple sclerosis (Becher and Segal, 2011) and systemic lupus erythematosus (Shin et al., 2011). It is therefore therapeutically desirable that immunosuppressive agents not only obstruct the initiation of antigen response but also achieve functional anergy by blocking the production of these pro-inflammatory cytokines from full-fledged effector CD4+ T cells (Getts et al.,

2011). To assess whether subglutinol A affects cytokine production by fully- differentiated effector CD4+ T cells, naïve LLO118 CD4+ T cells were first polarized in vitro under Th1 or Th17 condition for 4 days, during which they acquired IFNγ- and IL-

17-competence. In the 48 hours that followed, these effector Th1 and Th17 cells were then treated with various doses of subglutinol A, CsA or the vehicle control DMSO. For fully-differentiated Th1 cells, 10 nM of subglutinol A effectively diminished the percentage of IFNγ-producing cells from 90% to less than 1% (Figure 50A). A similar level of suppression was achieved with 100 nM of CsA. In terms of Th17 effector responses in the 10 nM range, subglutinol A suppressed IL-17A production less effectively than CsA. However, both drugs exhibited comparable potency at the higher doses of 50 nM and 100 nM (Figure 50B). Taken together, our data demonstrate that subglutinol A not only restricts the accumulation of auto-reactive T cells through its

160

expansion-inhibitory effect, but also dramatically attenuates inflammatory cytokine production by any remaining cells that survive.

161

Figure 50: Subglutinol A abolishes inflammatory cytokine production in fully differentiated Th1 and Th17 cells.

LLO CD4+ T cells were activated in vitro and polarized into fully differentiated (A) IFNγ-producing Th1 or (B) IL-17A-producing Th17 cells for the first 4 days in the absence of drugs. This was followed by 2 additional days of culture in the indicated concentrations of 1, CsA, or DMSO vehicle control (0 nM drugs). The percentages of cytokine-producing CD4+ T cells were then determined by intracellular staining and flow cytometry. Data shown is representative of two independent experiments.

162

5.2.4 Subglutinol A preferentially exacerbates mitochondrial depolarization in effector Th1 and Th17 cells

We next sought to elucidate the mode of action by which subglutinol A inhibits T cell activation. It is now increasingly appreciated that changes in T cell metabolism are necessary to fuel the functional and phenotypic changes associated with T cell activation.

When a naïve T cell is activated, it undergoes a “metabolic switch” from quiescent catabolism to active anabolism, giving rise to key biosynthetic substrates critical for T cell clonal expansion and cytokine production (Fox et al., 2005; Gerriets and Rathmell,

2012; Sena et al., 2013). Mitochondrion, the site of ATP-production by the mitochondrial electron transport chain, and the source of biosynthetic precursors by the Krebs cycle, is at the heart of this critical metabolic switch (Fernandez and Perl, 2009; Sena et al., 2013; van der Windt et al., 2012). In addition, the loss of mitochondrial integrity can also initiate the intrinsic apoptosis pathway to trigger cell death (Gatzka and Walsh, 2007;

Grimaldi et al., 2005; Gupta and Gollapudi, 2007). Based on the phenotype of subglutinol

A-treated T cells – impaired proliferation, increased cell death and reduced cytokine production – we hypothesized that subglutinol A may undermine T cell responses by compromising mitochondrial function. To test this hypothesis in fully-differentiated Th1 and Th17 cells, the number of mitochondria was evaluated by MitoTracker Deep Red FM labeling for mitochondrial mass, while mitochondrial integrity was evaluated by

MitoTracker Orange CMTMRos labeling for mitochondrial membrane potential. The total mitochondrial mass of fully differentiated Th1 was modestly reduced, while that of

Th17 cells was unchanged across the doses of subglutinol A tested, suggesting that

163

subglutinol A has little impact on mitochondrial replication or abundance in effector

CD4+ T cells (Figure 51A). However, in both Th1 and Th17 cells, treatment with subglutinol A induced mitochondrial depolarization in a dose-dependent manner, indicating that subglutinol A disrupts mitochondrial membrane integrity (Figure 51B).

On the contrary, CsA has previously been reported to inhibit mitochondrial permeability transition, thereby preventing mitochondrial depolarization (Liu and Murphy, 2009;

Nieminen et al., 1996). Therefore, subglutinol A may suppress inflammation though a distinct mechanism from CsA.

Mitochondria are essential organelles for all eukaryotic cells, and widespread mitochondrial dysfunction in normal tissues can result in severe adverse effects. To evaluate its target-cell specificity and potential as a therapeutic agent for auto- inflammation, we further interrogated whether subglutinol A similarly affects the mitochondria of other cell types. To this end, we examined both resting and activated antigen-presenting cells of the innate immune system, as exemplified by immature bone marrow-derived dendritic cells (imBMDCs) and mature BMDCs (mBMDCs), respectively. Additionally, we also examined the impact on normal stromal cells, as exemplified by the 3T3 fibroblast cell line. Interestingly, 100 nM of subglutinol A – the highest dose tested on activated T cells – altered neither mitochondrial mass nor depolarization in imBMDCs (Figure 52A; top panel), mBMDCs (Figure 52A; bottom panel), and 3T3 fibroblasts (Figure 52B). We postulate that activated T cells may have a heightened sensitivity to subglutinol A -induced mitochondrial dysfunction than other cell types, a premise that warrants further dose-escalation studies in the future. Taken

164

together, our data indicate that subglutinol A preferentially exacerbates mitochondrial damage in T cells, which may account for their impaired proliferation, death and blunted cytokine production.

165

Figure 51: Subglutinol A exacerbates mitochondrial depolarization in fully differentiated Th1 and Th17 cells in vitro.

Naı̈ve LLO CD4+ T cells were polarized in vitro under Th1- or Th17-skewing conditions for 4 days, followed by 2 additional days of culture in the indicated concentrations of subglutinol A or DMSO vehicle control (0 nM). Th1 and Th17 cells were then labeled with (A) MitoTracker Deep Red FM to assess mitochondrial mass and (B) MitoTracker Orange CMTMRos to assess mitochondrial membrane depolarization in live CD4+ T cells by flow cytometry. Numbers in histograms indicate mean fluorescence intensity of MitoTracker Deep Red FM staining. Data shown is representative of two independent experiments.

166

Figure 52: Subglutinol A does not affect mitochondrial mass or polarization in dendritic cells and stromal cells in vitro.

(A) Immature and LPS-matured bone marrow-derived dendritic cells (BMDCs) generated from mice were treated for 1 day with subglutinol A (100 nM), CsA (100 nM), or DMSO vehicle control. CD11c+ BMDCs were analyzed for mitochondria mass and polarization by flow cytometry. (B) 3T3 mouse fibroblasts were treated for 2 days and analyzed for mitochondria mass and polarization by flow cytometry. Numbers in histograms indicate mean fluorescence intensity of MitoTracker Deep Red FM and MitoTracker Orange CMTMRos staining. Data shown is representative of two independent experiments.

167

5.2.5 Subglutinol A treatment suppresses Th1-driven delayed-type hypersensitivity (DTH) reactions in vivo

To interrogate the immune-suppressive effects of subglutinol A in vivo, we utilized the classical Th1-driven inflammatory response of delayed-type hypersensitivity

(DTH) (Kobayashi et al., 2001). DTH is an antigen-specific inflammatory reaction in the skin that is triggered by repeated exposure to certain antigens, resulting in the activation and infiltration of skin-homing antigen-specific T cells (Saint-Mezard et al., 2004;

Vocanson et al., 2009). The induction of the DTH response involves 2 stages: a sensitization phase and an effector phase. When T cells that have been previously sensitized by an antigen re-encounter the same antigen underneath the skin, the Th1- dominated effector response ensues, resulting in the manifestation of DTH. DTH is characterized by profound lymphocytic recruitment, Th1 CD4+ T cell-mediated cytokine

(e.g., IFNγ) secretion, tissue damage and local swelling at the site of antigenic stimulation.

We utilized a mouse model of DTH induced by the highly-immunogenic protein, keyhole limpet hemocyanin (KLH) (Allen, 2013). C57BL/6J mice were first sensitized subcutaneously by introducing a 100 μg of the KLH antigen emulsified in complete

Freund’s adjuvant (CFA). 7 days after the initial sensitization, each mouse was re- exposed on one footpad to 50 μg KLH and simultaneously treated with 16 nmol (0.273 mg/kg) of subglutinol A, 16 nmol (0.769 mg/kg) of CsA, or the DMSO vehicle control.

As a negative control, PBS alone (antigen-free) was also injected into the other

168

previously sensitized footpad. 48 hours after antigenic re-exposure, we assessed swelling and lymphocytic infiltration into the footpads.

As compared to the footpad without re-exposure (PBS) that contained a few lymphocytes and maintained normal tissue integrity, mice that were re-exposed to KLH in combination with the DMSO vehicle control (KLH+DMSO) exhibited extensive lymphocytic infiltration into the dermis and severe disruption of tissue architecture.

Consistent with earlier reports (Bussiere et al., 1991; Remitz et al., 1996; Tchervenkov et al., 1985), CsA did not alleviate DTH-induced footpad swelling at low doses (Figure 53); this was attributed to effects of CsA on non-T cells, which may counterproductively promote skin inflammation (Remitz et al., 1996). However, treatment with equimolar amounts of subglutinol A (KLH+SubA) significantly blocked lymphocyte accumulation and ameliorated tissue injury (Figure 53). In addition, subglutinol A also significantly reduced inflammation-induced footpad swelling. These data not only corroborate our findings from the in vitro assays, but more importantly demonstrate the in vivo immune- suppressive potency, efficacy and target-cell specificity of subglutinol A at low drug doses.

169

Figure 53: Subglutinol A treatment suppresses antigen-induced DTH responses in vivo.

C57BL/6J mice previously sensitized to the KLH antigen had their left footpads re- exposed to the KLH antigen and were treated with either 16 nmol (0.273 mg/kg) of subglutinol A (KLH + SubA), 16 nmol (0.769 mg/kg) of CsA (KLH + CsA), or DMSO vehicle control (KLH + DMSO). As a control for background swelling, their right footpads were not re-exposed to KLH, but were injected with an equal volume of PBS (PBS). Two days after re-exposure, footpad swelling was read, mice were sacrificed, and their footpads excised for histological analysis by hematoxylin and eosin staining where indicated. Images of histological sections are representative of 5 mice per group; p-value is determined by the two-tailed unpaired t-test.

170

5.3 Discussion

A comprehensive evaluation of the capacity of subglutinol A, a novel immunosuppressive natural product, was performed to demonstrate its potential as a novel therapeutic agent for autoimmune diseases. Subglutinol A profoundly inhibited T cell proliferation, survival, and pathogenic cytokine production by fully differentiated

Th1 and Th17 cells in vitro, potentially by aggravating mitochondrial damage in T cells.

Importantly, immunosuppressive doses of subglutinol A did not affect mitochondrial integrity in non-T cells, such as antigen-presenting BMDCs and stromal fibroblasts.

Moreover, low-dose therapy with subglutinol A, but not CsA, was efficacious at attenuating the Th1-driven DTH response, demonstrating its anti-inflammatory, immunosuppressive efficacy and target-cell specificity in vivo. These results suggest that subglutinol A may provide opportunities for new, innovative, and efficacious therapies to treat autoimmune diseases, as well as better post-transplantation cares. Currently, structure-function relationship studies of subglutinol A aimed at improving its efficacy and safety and further studies to identify the mode of action of subglutinol A are in progress.

171

6. General discussion and future perspectives

The immune system’s unique ability to distinguish between self and non-self is a double-edged sword. To ensure immune-homeostasis, intricate mechanisms are at play to enforce tolerance towards self while potentiating attack against foreign antigens.

However, when these protective mechanisms are hijacked or go awry, the resulting immune-dysregulation becomes deleterious to the host, implicating inflammation as an etiological factor underlying a wide array of human diseases. The central roles of T cells executioners and coordinators of immune responses, coupled to their amenability to functional manipulation, have made them critical targets for the development of immune- modulatory therapies for cancer and autoimmunity. Although research on T cell-based immune-modulation has achieved significant milestones, broad clinical application of these approaches remain limited by (i) their susceptibility to immune-suppression in vivo,

(ii) controversy over the “ideal” protective T cell subset and (iii) the potential for severe adverse effects and toxicity. This dissertation contributes to advancing the field of T cell- targeted immunotherapy, by identifying novel entities for modulating the effector functions of T cells and bringing to attention important considerations in the design of successful therapeutic strategies.

In Chapter 3, we identified and mechanistically dissected a miRNA-targeting approach that potently modulates CD8+ T cell effector responses by instilling resilience to tumor-induced functional-suppression. We additionally established a role for miR-23a in the immune-pathogenesis of human cancers, underscoring the translational relevance of this miRNA target. Importantly, the miR-23a decoy developed in these studies can 172

serve as a potent gene therapy tool that may feasibly be incorporated with current CAR T cell and high-affinity TCR engineering approaches: the short length of the miR-23a decoy moiety (~200 basepairs) allows it to be introduced easily into dual-promoter coexpression viral vectors, enabling the simultaneous inhibition of miR-23a in tumor- redirected CD8+ T cells for ACT.

Targeting miR-23a in tumor-specific CD8+ T cells has important implications for

ACT. Preclinical studies on solid tumor immunotherapy have demonstrated that despite exhibiting superior anti-tumor potency in vitro, T cells engineered with optimized high- affinity TCRs and CARs still become incapacitated in vivo (Moon et al., 2014). That strong T cell activation signals alone are inadequate at overcoming immune barriers within the tumor microenvironment underscores the need for additional enhancements to

ACT. In Chapter 4, we evaluated the utility of targeting miR-23a in CD8+ CAR T cells as one potential approach. We provided evidence that engineered CAR T cells remain vulnerable to immune-suppression by the TGFβ-miR-23a pathway (Figure 41), supporting the relevance of targeting miR-23 for CAR-based ACT. The anti-tumor efficacy of miR-23a-deficient CARs, though, was severely crippled by their failure to persist in vivo. Although further optimization of treatment protocols is required, we provided the proof-of-concept that miRNA-based approaches represent compelling and viable alternatives for enhancing the immune-competence of T cell products used in

ACT. This dissertation therefore serves as a jumping-off point for further target discovery studies, by highlighting the need to identify additional T cell-modulatory miRNAs capable of conferring immune-competence within the tumor microenvironment.

173

From a mechanistic standpoint, our studies on miR-23a echo the ongoing conundrum on the T cell subset that is most therapeutically suited for ACT. As discussed in Chapter 1.2.3, transcriptional programs required for highly-cytotoxic effector phenotypes are mutually antagonistic with those necessary for differentiation to long- lived CD8+ memory cells. This raises debate over whether terminally-differentiated

CD8+ T cells fully-equipped with immediate cytotoxic functions, or their poorly- cytotoxic but multipotent counterparts, are more efficacious populations for ACT.

Supporters of the former have proven that IL-2 driven CD8+ SLECs and Tem mediate potent protection due to their ability to survey peripheral tissues and execute immediate cytotoxicity (Kilinc et al., 2009; Miller et al., 2008; Olson et al., 2013; Radvanyi et al.,

2012). Proponents of the latter, on the other hand, have demonstrated the superior ability of IL-7- and IL-15-driven CD8+ memory stem cells (Tscm) and Tcm to survive long- term and retain the capacity to differentiate into and reconstitute effector cells (Berger et al., 2008; Cieri et al., 2013; Gattinoni et al., 2011; Graef et al., 2014; Kaneko et al.,

2009). Blimp-1, the direct target of miR-23a, instructs CD8+ T cell differentiation into functionally-robust SLECs and Tem cells under immune-activating conditions (Kallies et al., 2009; Rutishauser et al., 2009), and promotes CD8+ T cell exhaustion during chronic inflammation (Shin et al., 2009); we therefore anticipated ACT involving permanent miR-23a inhibition to be met with substantial skepticism. Indeed, although miR-23a- deficient CAR T cells were strongly potent and functionally resilient in vitro, our preliminary efficacy tests in lymphoreplete tumor-bearing mice revealed that therapy with miR-23a-abrogated CAR T cells was ineffective due to their failure to persist in

174

vivo. Within the confines of our studies, this reiterates the need for supportive preconditioning or cytokine regimens should miR-23a-targeting be integrated into ACT.

In a broader perspective, our findings imply ACT may be most effective with a heterogeneous and multi-functional, but defined composition of T cell subsets at various differentiation states. To date, T cell products clinically manufactured for ACT comprise a bulk mixture of CD4+ and CD8+ T cells selected solely based on tumor-reactivity, with no defined standards specifying CD4+:CD8+ ratios or differentiation statuses of reinfused T cells. These are important considerations, given that CD4+ T cell “help” enhances the quality, nature and maintenance of CD8+ T cell responses, while a mixture of highly-cytotoxic effectors and multipotent memory cells confer immediate and long- lasting anti-tumor protection. Furthermore, due to differences in their functional properties and anatomical localization, the types of memory T cell subsets generated may have far-reaching impact on the durability of clinical responses. In particular, emerging evidence has highlighted that CD8+ Trm’s residing in peripheral tissues afford superior cell-mediated immunity against virus reinfection; moreover, increased numbers of tumor- infiltrating CD8+ Trm’s are associated with improved prognosis and prolonged overall survival among ovarian cancer patients (Webb et al., 2014). The anatomical localization of CD8+ Trms to peripheral sites poise them as tissue-resident sentinels for effective anti- tumor immune-surveillance, making them a compelling memory population for use in

ACT. As a result, future efforts in developing protective T cell compositions for ACT should focus on enhancing the representation of CD8+ Trm’s. The current lack of uniformity between T cell products derived from different patients may account for the

175

large inter-individual variation in clinical outcomes to ACT. Optimizing and defining subset compositions will ensure the quality of reinfused T cell products, thereby providing a means to improving efficacy and consistency in clinical outcomes of ACT.

Although our studies have focused mainly on the function and fate of the adoptively-transferred CAR T cells, further mechanistic studies are required to elucidate how CAR therapy influences the endogenous host immune system in the long-run. Prior therapy with CAR T cells prevents the recurrence of tumors lacking expression of the targeted antigen (Barber et al., 2008; Sampson et al., 2014; Spear et al., 2013a), demonstrating the immune-potentiating impact of CAR therapy on endogenous anti- tumor defense. These protective effects elicited have been attributed to CAR T cell- induced mobilization of endogenous immunity within the tumor microenvironment. In preclinical studies utilizing mouse models of ovarian cancer, the adoptively transferred tumor-infiltrating NKG2D-CAR T cells were shown alter the tumor microenvironment by secreting pro-inflammatory cytokines, such as GM-CSF and IFNγ. These in turn created an immune-stimulatory milieu within the tumor microenvironment, facilitating enhanced intratumoral innate immune cell activation and antigen-presentation, increased effector NK cells and T cells recruitment into tumors, as well as reduced tumor- infiltrating Treg numbers (Barber et al., 2009; Spear et al., 2012). Abalation of macrophages by clodronate-containing liposomes prior to ACT abrogated the efficacy of

NKG2D-CAR therapy, identifying NKG2D-CAR-induced macrophage activation as the essential initiating step responsible for perpetuating downstream mobilization of endogenous immunity during the response to primary tumors (Spear et al., 2012).

176

However, the long-term protection against rechallenge with secondary tumors lacking the targeted antigen cannot be explained merely by innate immune activation; clearly, these must be mediated by the development of a broad-spectrum immunological memory against additional tumor antigens. Although secondary protection is often attributed to the induction of endogenous anti-tumor memory T cells (Barber et al.,

2008), three important aspects that remain ill-defined and warrant further characterization are:

(i) The subsets of endogenous tumor-reactive memory T cells generated following

CAR therapy.

As discussed in Chapter 1.2.1, the differential homing and functional capacities

of CD8+ Tcm, Tem and Trm memory subsets dictate their efficiencies in

surveying for and eradicating recurring tumor cells. Understanding how CAR

therapy affects memory T cell differentiation may shed light on novel

approaches aimed at prolonging the durability of ACT. For instance,

optimizing CAR therapy to favor the development of tumor tissue-resident

Trm’s may maximize local immune-surveillance at primary tumor sites,

whereas the generation of recirculating Tcm’s and Tem’s may more effectively

target metastatic lesions at secondary sites.

(ii) How the quality and quantity of endogenous tumor-reactive memory T cells

correlates with prognosis following CAR therapy.

In ACT, the in vivo persistence of the adoptively transferred tumor-reactive T

cells correlates positively with patient overall survival (Rosenberg et al., 2011).

177

Moreover, the density of tumor-infiltrating endogenous T cells serves as a

more accurate prognostic predictor of cancer recurrence than the conventional

cancer staging system (Galon et al., 2006; Mlecnik et al., 2011). Therefore,

following CAR therapy, the persistence and function of endogenous tumor-

reactive T cells may serve as an immune-monitoring parameter to predict

clinical responses to CAR-based ACT and the risk of relapse. To this end, T

cells isolated from patient PBMCs may be cultured with autologous tumor

cells lacking the targeted antigen (e.g. by shRNA-mediated knockdown or by

cell sorting of antigen-negative tumor variants). The abundance of tumor-

reactive endogenous memory T cell abundance may be determined by the IFNγ

ELISPOT assay, their functional capacity by flow cytometric analysis of IFNγ

expression.

(iii) The breath of endogenous tumor-reactive T cell antigen specificities elicited by

CAR therapy and its prognostic value.

Since CAR therapy lowers the threshold for the activation of endogenous

tumor-reactive T cells, it is reasonable to speculate that the resultant broadened

spectrum of anti-tumor TCR repertoire should confer enhanced protection

against tumor relapse. However, how CAR therapy affects the endogenous

TCR repertoire diversity, and how this in turn associates with tumor

progression, has yet to be characterized; this area is therefore in need of

detailed investigation through preclinical studies. If endogenous T cell epitope

spread following CAR therapy correlates positively with tumor regression, it

178

will provide a strong impetus for future research to delve into widening the

breadth of endogenous anti-tumor T cell reactivity by enhancing CAR-elicited

epitope spread.

In spite of the enthusiasm surrounding ACT, the risk of severe adverse reactions emphasizes a need for caution. In TIL-based ACT for melanoma, tumor regression is often accompanied by autoimmune vitiligo (Dudley et al., 2002; Overwijk et al., 2003).

In recent clinical trials of CAR therapy, cytokine storms and massive tissue damage have led to serious adverse events, and even patient deaths (Brentjens et al., 2010; Lamers et al., 2006; Morgan et al., 2010).

Based on our current understanding on the interplay between reinfused CAR T cells and endogenous immunity, we have just begun to scratch the surface of the resounding impact CAR therapy may potentially have not only on anti-tumor, but also anti-self, responses. The immune-activating milieu generated by CAR therapy is postulated to relieve microenvironmental-induced immune-suppression, thereby mobilizing endogenous anti-tumor T cells (Spear et al., 2012). Indeed, the induction of such broad-spectrum anti-tumor T cells extends the clinical application of targeted-CAR therapy to include antigen-heterogeneous tumors. Nevertheless, this also raises concerns that T cells recognizing TAAs and normal tissue antigens may be activated or form long- lived memory cells, resulting in “on-tumor, off-target” effects that culminate in cytokine storms or autoimmune disorders. The risks of CAR-induced auto-inflammation may run high especially in poorly-immunogenic tumors such as GBM, which are antigenically similar to normal cells and express few tumor-specific neo-antigens (Vogelstein et al.,

179

2013). In our preclinical studies on GBM immunotherapy, we demonstrated that even when EGFRvIII-CAR T cells were delivered directly into the immune-privileged brain, they were not completely anatomically confined by the BBB: a portion of CAR T cells still escaped the BBB into the periphery (Figure 44). The leakiness of the BBB induced by the tumor-bearing state suggests that intratumorally-activated endogenous T cells are equally likely to exit into the periphery, where they may cross-react with antigens expressed by normal cells.

To increase the safety profile of ACT, a safety switch may be incorporated into the infused T cell products, enabling their selective depletion for the timely management of adverse effects. For instance, during ex vivo expansion, tumor-specific T cells may be genetically modified by viral transduction of suicide genes, such as the herpes simplex virus thymidine kinase (HSV-TK) (Bonini et al., 1997; Ciceri et al., 2009; Hsu et al.,

2013) or inducible caspase-9 (iCasp9) (Di Stasi et al., 2011). Alternatively, these suicide genes may be directly integrated into CAR vectors to allow their co-expression (Budde et al., 2013). Upon the onset of adverse events, the apoptosis of HSV-TK- and iCasp9- modified T cells can be triggered by pharmacological administration of ganciclovir

(Bonini et al., 1997) or the dimerizing drug AP1903 (Di Stasi et al., 2011). The iCasp9 suicide gene, though, is the preferred choice: unlike HSV-TK whose activity is restricted to diving cells and requires at least days to take effect (Ciceri et al., 2009), iCasp9 can eliminate as much as 90% of proliferating and quiescent targets within 30 minutes (Di

Stasi et al., 2011). Nonetheless, it should be noted that suicide genes can only directly restrain the reinfused T cells, and have no bearing on the endogenous immune activation

180

that may have already been set in motion. Understanding how CAR therapy shapes the magnitude, kinetics and distribution of host immune cells will represent the first step in pre-empting and devising management strategies to preferentially limit endogenously- arising auto-inflammation, while sparing anti-tumor immune responses.

In conclusion, this dissertation has furthered the development of T cell-based approaches for immune-modulation, by identifying novel, viable and clinically-relevant targets that can be feasibly integrated into current therapeutic procedures. Importantly, our studies call special attention to the appeal of miRNAs as a class of effective immune- modulatory elements whose function can be feasibly and permanently altered to manipulate CD8+ T cell cytotoxicity. In two types of CD8+ T cells commonly used in

ACT – unmodified tumor-specific CD8+ T cells and CAR-modified CD8+ cells – miR-

23a abrogation enhanced their anti-tumor effector function and immune-competence, making miR-23a a compelling gene therapy target for improving the efficacy of ACT.

Currently, further studies are underway to define the utility of targeting miR-23a for ACT in the long-run. In addition to miRNAs, our studies presented an alternative means of T cell immune-modulation through the discovery and characterization of novel pharmacological agents, such as subglutinol A. Identifying the drug target and mode of action of subglutinol A will be crucial to guide future structure-function optimization, facilitating the development of analogues with heightened T cell specificities and safety profiles.

181

References

Agata, Y., Kawasaki, A., Nishimura, H., Ishida, Y., Tsubata, T., Yagita, H., and Honjo, T. (1996). Expression of the PD-1 antigen on the surface of stimulated mouse T and B lymphocytes. International immunology 8, 765-772.

Ahmadzadeh, M., and Rosenberg, S.A. (2005). TGF-beta 1 attenuates the acquisition and expression of effector function by tumor antigen-specific human memory CD8 T cells. J Immunol 174 , 5215-5223.

Ahmadzadeh M, R.S. (2005). TGF-beta 1 attenuates the acquisition and expression of effector function by tumor antigen-specific human memory CD8 T cells. J Immunol 174 , 5215-5223.

Ahmed, K.A., Wang, L., and Xiang, J. (2012). A new dynamic model of three cell interactions for CTL responses. Oncoimmunology 1, 1430-1432.

Ahmed, N., Salsman, V.S., Kew, Y., Shaffer, D., Powell, S., Zhang, Y.J., Grossman, R.G., Heslop, H.E., and Gottschalk, S. (2010). HER2-specific T cells target primary glioblastoma stem cells and induce regression of autologous experimental tumors. Clin Cancer Res 16 , 474-485.

Al-Nedawi, K., Meehan, B., Micallef, J., Lhotak, V., May, L., Guha, A., and Rak, J. (2008). Intercellular transfer of the oncogenic receptor EGFRvIII by microvesicles derived from tumour cells. Nature cell biology 10 , 619-624.

Alderson, K.L., and Sondel, P.M. (2011). Clinical cancer therapy by NK cells via antibody-dependent cell-mediated cytotoxicity. Journal of biomedicine & biotechnology 2011 , 379123.

Alizadeh, D., Katsanis, E., and Larmonier, N. (2013). The multifaceted role of Th17 lymphocytes and their associated cytokines in cancer. Clin Dev Immunol 2013 , 957878.

Allen, I.C. (2013). Delayed-type hypersensitivity models in mice. Methods Mol Biol 1031 , 101-107.

182

Alves, N.L., van Leeuwen, E.M., Derks, I.A., and van Lier, R.A. (2008). Differential regulation of human IL-7 receptor alpha expression by IL-7 and TCR signaling. J Immunol 180 , 5201-5210.

Andreae, S., Piras, F., Burdin, N., and Triebel, F. (2002). Maturation and activation of dendritic cells induced by lymphocyte activation gene-3 (CD223). J Immunol 168 , 3874- 3880.

Annunziato, F., Manetti, R., Tomasevic, I., Guidizi, M.G., Biagiotti, R., Gianno, V., Germano, P., Mavilia, C., Maggi, E., and Romagnani, S. (1996). Expression and release of LAG-3-encoded protein by human CD4+ T cells are associated with IFN-gamma production. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 10 , 769-776.

Apetoh, L., Ghiringhelli, F., Tesniere, A., Criollo, A., Ortiz, C., Lidereau, R., Mariette, C., Chaput, N., Mira, J.P., Delaloge, S. , et al. (2007). The interaction between HMGB1 and TLR4 dictates the outcome of anticancer chemotherapy and radiotherapy. Immunological reviews 220 , 47-59.

Araki, K., Turner, A.P., Shaffer, V.O., Gangappa, S., Keller, S.A., Bachmann, M.F., Larsen, C.P., and Ahmed, R. (2009). mTOR regulates memory CD8 T-cell differentiation. Nature 460 , 108-112.

Ariotti, S., Hogenbirk, M.A., Dijkgraaf, F.E., Visser, L.L., Hoekstra, M.E., Song, J.Y., Jacobs, H., Haanen, J.B., and Schumacher, T.N. (2014). T cell memory. Skin-resident memory CD8(+) T cells trigger a state of tissue-wide pathogen alert. Science 346 , 101- 105.

Atanackovic, D., Block, A., de Weerth, A., Faltz, C., Hossfeld, D.K., and Hegewisch- Becker, S. (2004). Characterization of effusion-infiltrating T cells: benign versus malignant effusions. Clin Cancer Res 10 , 2600-2608.

Balague, C., Kunkel, S.L., and Godessart, N. (2009). Understanding autoimmune disease: new targets for drug discovery. Drug Discov Today 14 , 926-934.

Baltimore D, B.M., O'Connell RM, Rao DS, Taganov KD. (2008). MicroRNAs: new regulators of immune cell development and function. Nat Immunol 9, 839-845.

183

Bandukwala, H.S., Gagnon, J., Togher, S., Greenbaum, J.A., Lamperti, E.D., Parr, N.J., Molesworth, A.M., Smithers, N., Lee, K., Witherington, J. , et al. (2012). Selective inhibition of CD4+ T-cell cytokine production and autoimmunity by BET protein and c- Myc inhibitors. Proc Natl Acad Sci U S A 109 , 14532-14537.

Banerjee, A., Gordon, S.M., Intlekofer, A.M., Paley, M.A., Mooney, E.C., Lindsten, T., Wherry, E.J., and Reiner, S.L. (2010). Cutting edge: The transcription factor eomesodermin enables CD8+ T cells to compete for the memory cell niche. J Immunol 185 , 4988-4992.

Baragano Raneros, A., Martin-Palanco, V., Fernandez, A.F., Rodriguez, R.M., Fraga, M.F., Lopez-Larrea, C., and Suarez-Alvarez, B. (2014). Methylation of NKG2D ligands contributes to immune system evasion in acute myeloid leukemia. Genes and immunity 0.

Baran, K., Dunstone, M., Chia, J., Ciccone, A., Browne, K.A., Clarke, C.J., Lukoyanova, N., Saibil, H., Whisstock, J.C., Voskoboinik, I., and Trapani, J.A. (2009). The molecular basis for perforin oligomerization and transmembrane pore assembly. Immunity 30 , 684- 695.

Barber, A., Rynda, A., and Sentman, C.L. (2009). Chimeric NKG2D expressing T cells eliminate immunosuppression and activate immunity within the ovarian tumor microenvironment. J Immunol 183 , 6939-6947.

Barber, A., Zhang, T., and Sentman, C.L. (2008). Immunotherapy with chimeric NKG2D receptors leads to long-term tumor-free survival and development of host antitumor immunity in murine ovarian cancer. J Immunol 180 , 72-78.

Barry, M., Heibein, J.A., Pinkoski, M.J., Lee, S.F., Moyer, R.W., Green, D.R., and Bleackley, R.C. (2000). Granzyme B short-circuits the need for caspase 8 activity during granule-mediated cytotoxic T-lymphocyte killing by directly cleaving Bid. Mol Cell Biol 20 , 3781-3794.

Bartel, D.P. (2009). MicroRNAs: target recognition and regulatory functions. Cell 136 , 215-233.

Bauer, S., Groh, V., Wu, J., Steinle, A., Phillips, J.H., Lanier, L.L., and Spies, T. (1999). Activation of NK cells and T cells by NKG2D, a receptor for stress-inducible MICA. Science 285 , 727-729. 184

Beal, A.M., Anikeeva, N., Varma, R., Cameron, T.O., Vasiliver-Shamis, G., Norris, P.J., Dustin, M.L., and Sykulev, Y. (2009). Kinetics of early T cell receptor signaling regulate the pathway of lytic granule delivery to the secretory domain. Immunity 31 , 632-642.

Becher, B., and Segal, B.M. (2011). T(H)17 cytokines in autoimmune neuro- inflammation. Curr Opin Immunol 23 , 707-712.

Bedoui, S., Whitney, P.G., Waithman, J., Eidsmo, L., Wakim, L., Caminschi, I., Allan, R.S., Wojtasiak, M., Shortman, K., Carbone, F.R. , et al. (2009). Cross-presentation of viral and self antigens by skin-derived CD103+ dendritic cells. Nat Immunol 10 , 488- 495.

Bell, A.C., West, A.G., and Felsenfeld, G. (1999). The protein CTCF is required for the enhancer blocking activity of vertebrate insulators. Cell 98 , 387-396.

Bennicelli JL, G.D.t. (1993). Production of multiple cytokines by cultured human melanomas. Exp Dermatol 2, 186-190.

Bennicelli, J.L., and Guerry, D.t. (1993). Production of multiple cytokines by cultured human melanomas. Exp Dermatol 2, 186-190.

Bergamaschi, C., Rosati, M., Jalah, R., Valentin, A., Kulkarni, V., Alicea, C., Zhang, G.M., Patel, V., Felber, B.K., and Pavlakis, G.N. (2008). Intracellular interaction of interleukin-15 with its receptor alpha during production leads to mutual stabilization and increased bioactivity. J Biol Chem 283 , 4189-4199.

Berger, C., Berger, M., Hackman, R.C., Gough, M., Elliott, C., Jensen, M.C., and Riddell, S.R. (2009). Safety and immunologic effects of IL-15 administration in nonhuman primates. Blood 114 , 2417-2426.

Berger, C., Jensen, M.C., Lansdorp, P.M., Gough, M., Elliott, C., and Riddell, S.R. (2008). Adoptive transfer of effector CD8+ T cells derived from central memory cells establishes persistent T cell memory in primates. J Clin Invest 118 , 294-305.

Biburger, M., Lux, A., and Nimmerjahn, F. (2014). How immunoglobulin G antibodies kill target cells: revisiting an old paradigm. Advances in immunology 124 , 67-94.

185

Bladergroen, B.A., Meijer, C.J., ten Berge, R.L., Hack, C.E., Muris, J.J., Dukers, D.F., Chott, A., Kazama, Y., Oudejans, J.J., van Berkum, O., and Kummer, J.A. (2002). Expression of the granzyme B inhibitor, protease inhibitor 9, by tumor cells in patients with non-Hodgkin and Hodgkin lymphoma: a novel protective mechanism for tumor cells to circumvent the immune system? Blood 99 , 232-237.

Blank, C.U., and Enk, A. (2014). Therapeutic use of anti-CTLA-4 antibodies. International immunology.

Blankenstein, T., Coulie, P.G., Gilboa, E., and Jaffee, E.M. (2012). The determinants of tumour immunogenicity. Nature Reviews Cancer 12 , 307-313.

Bodmer, S., Strommer, K., Frei, K., Siepl, C., de Tribolet, N., Heid, I., and Fontana, A. (1989). Immunosuppression and transforming growth factor-beta in glioblastoma. Preferential production of transforming growth factor-beta 2. J Immunol 143 , 3222-3229.

Boissier, M.C., Semerano, L., Challal, S., Saidenberg-Kermanac'h, N., and Falgarone, G. (2012). Rheumatoid arthritis: from autoimmunity to synovitis and joint destruction. J Autoimmun 39 , 222-228.

Bonini, C., Ferrari, G., Verzeletti, S., Servida, P., Zappone, E., Ruggieri, L., Ponzoni, M., Rossini, S., Mavilio, F., Traversari, C., and Bordignon, C. (1997). HSV-TK gene transfer into donor lymphocytes for control of allogeneic graft-versus-leukemia. Science 276 , 1719-1724.

Boockvar, J.A., Kapitonov, D., Kapoor, G., Schouten, J., Counelis, G.J., Bogler, O., Snyder, E.Y., McIntosh, T.K., and O'Rourke, D.M. (2003). Constitutive EGFR signaling confers a motile phenotype to neural stem cells. Molecular and cellular neurosciences 24 , 1116-1130.

Bossard, C., Belhadj, K., Reyes, F., Martin-Garcia, N., Berger, F., Kummer, J.A., Briere, J., Baglin, A.C., Cheze, S., Bosq, J. , et al. (2007). Expression of the granzyme B inhibitor PI9 predicts outcome in nasal NK/T-cell lymphoma: results of a Western series of 48 patients treated with first-line polychemotherapy within the Groupe d'Etude des Lymphomes de l'Adulte (GELA) trials. Blood 109 , 2183-2189.

Boyman, O., and Sprent, J. (2012). The role of interleukin-2 during homeostasis and activation of the immune system. Nat Rev Immunol 12 , 180-190. 186

Brabletz, T., Pfeuffer, I., Schorr, E., Siebelt, F., Wirth, T., and Serfling, E. (1993). Transforming growth factor beta and cyclosporin A inhibit the inducible activity of the interleukin-2 gene in T cells through a noncanonical octamer-binding site. Mol Cell Biol 13 , 1155-1162.

Brahmer, J.R., Drake, C.G., Wollner, I., Powderly, J.D., Picus, J., Sharfman, W.H., Stankevich, E., Pons, A., Salay, T.M., McMiller, T.L. , et al. (2010). Phase I study of single-agent anti-programmed death-1 (MDX-1106) in refractory solid tumors: safety, clinical activity, pharmacodynamics, and immunologic correlates. J Clin Oncol 28 , 3167- 3175.

Brahmer, J.R., Tykodi, S.S., Chow, L.Q., Hwu, W.J., Topalian, S.L., Hwu, P., Drake, C.G., Camacho, L.H., Kauh, J., Odunsi, K. , et al. (2012). Safety and activity of anti-PD- L1 antibody in patients with advanced cancer. N Engl J Med 366 , 2455-2465.

Brandt, K., Bulfone-Paus, S., Foster, D.C., and Ruckert, R. (2003). Interleukin-21 inhibits dendritic cell activation and maturation. Blood 102 , 4090-4098.

Brentjens, R., Yeh, R., Bernal, Y., Riviere, I., and Sadelain, M. (2010). Treatment of chronic lymphocytic leukemia with genetically targeted autologous T cells: case report of an unforeseen adverse event in a phase I clinical trial. Molecular therapy : the journal of the American Society of Gene Therapy 18 , 666-668.

Brentjens, R.J., Davila, M.L., Riviere, I., Park, J., Wang, X., Cowell, L.G., Bartido, S., Stefanski, J., Taylor, C., Olszewska, M. , et al. (2013). CD19-targeted T cells rapidly induce molecular remissions in adults with chemotherapy-refractory acute lymphoblastic leukemia. Sci Transl Med 5, 177ra138.

Brignone, C., Escudier, B., Grygar, C., Marcu, M., and Triebel, F. (2009). A phase I pharmacokinetic and biological correlative study of IMP321, a novel MHC class II agonist, in patients with advanced renal cell carcinoma. Clin Cancer Res 15 , 6225-6231.

Brignone, C., Gutierrez, M., Mefti, F., Brain, E., Jarcau, R., Cvitkovic, F., Bousetta, N., Medioni, J., Gligorov, J., Grygar, C. , et al. (2010). First-line chemoimmunotherapy in metastatic breast carcinoma: combination of paclitaxel and IMP321 (LAG-3Ig) enhances immune responses and antitumor activity. J Transl Med 8, 71.

187

Brincks, E.L., and Woodland, D.L. (2010). Novel roles for IL-15 in T cell survival. F1000 biology reports 2, 67.

Bromberg, J.F., Horvath, C.M., Wen, Z., Schreiber, R.D., and Darnell, J.E., Jr. (1996). Transcriptionally active Stat1 is required for the antiproliferative effects of both interferon alpha and interferon gamma. Proc Natl Acad Sci U S A 93 , 7673-7678.

Brucet, M., Marques, L., Sebastian, C., Lloberas, J., and Celada, A. (2004). Regulation of murine Tap1 and Lmp2 genes in macrophages by interferon gamma is mediated by STAT1 and IRF-1. Genes and immunity 5, 26-35.

Budde, L.E., Berger, C., Lin, Y., Wang, J., Lin, X., Frayo, S.E., Brouns, S.A., Spencer, D.M., Till, B.G., Jensen, M.C. , et al. (2013). Combining a CD20 chimeric antigen receptor and an inducible caspase 9 suicide switch to improve the efficacy and safety of T cell adoptive immunotherapy for lymphoma. PLoS One 8, e82742.

Burkett, P.R., Koka, R., Chien, M., Chai, S., Boone, D.L., and Ma, A. (2004). Coordinate expression and trans presentation of interleukin (IL)-15Ralpha and IL-15 supports natural killer cell and memory CD8+ T cell homeostasis. J Exp Med 200 , 825-834.

Bussiere, J.L., Mather, G.G., and Exon, J.H. (1991). Effect of cyclosporine on 3- methylcholanthrene-induced carcinogenesis and immune responses in the rat. Immunobiology 182 , 205-215.

Caminschi, I., Lahoud, M.H., and Shortman, K. (2009). Enhancing immune responses by targeting antigen to DC. Eur J Immunol 39 , 931-938.

Carlyle, J.R., Jamieson, A.M., Gasser, S., Clingan, C.S., Arase, H., and Raulet, D.H. (2004). Missing self-recognition of Ocil/Clr-b by inhibitory NKR-P1 natural killer cell receptors. Proc Natl Acad Sci U S A 101 , 3527-3532.

Carrington, J.C., and Ambros, V. (2003). Role of microRNAs in plant and animal development. Science 301 , 336-338.

Carrington JC, A.V. (2003). Role of microRNAs in plant and animal development. Science 301 , 336-338.

188

Carson, W.E., Ross, M.E., Baiocchi, R.A., Marien, M.J., Boiani, N., Grabstein, K., and Caligiuri, M.A. (1995). Endogenous production of interleukin 15 by activated human monocytes is critical for optimal production of interferon-gamma by natural killer cells in vitro. J Clin Invest 96 , 2578-2582.

Castiglioni, P., Gerloni, M., Cortez-Gonzalez, X., and Zanetti, M. (2005). CD8 T cell priming by B lymphocytes is CD4 help dependent. Eur J Immunol 35 , 1360-1370.

Castriconi, R., Cantoni, C., Della Chiesa, M., Vitale, M., Marcenaro, E., Conte, R., Biassoni, R., Bottino, C., Moretta, L., and Moretta, A. (2003). Transforming growth factor beta 1 inhibits expression of NKp30 and NKG2D receptors: consequences for the NK-mediated killing of dendritic cells. Proc Natl Acad Sci U S A 100 , 4120-4125.

Chen, B., He, L., Savell, V.H., Jenkins, J.J., and Parham, D.M. (2000). Inhibition of the interferon-gamma/signal transducers and activators of transcription (STAT) pathway by hypermethylation at a STAT-binding site in the p21WAF1 promoter region. Cancer Res 60 , 3290-3298.

Chen, C.H., Seguin-Devaux, C., Burke, N.A., Oriss, T.B., Watkins, S.C., Clipstone, N., and Ray, A. (2003). Transforming growth factor beta blocks Tec kinase phosphorylation, Ca2+ influx, and NFATc translocation causing inhibition of T cell differentiation. J Exp Med 197 , 1689-1699.

Chhabra, A. (2011). TCR-engineered, customized, antitumor T cells for cancer immunotherapy: advantages and limitations. ScientificWorldJournal 11 , 121-129.

Cho, O.H., Shin, H.M., Miele, L., Golde, T.E., Fauq, A., Minter, L.M., and Osborne, B.A. (2009). Notch regulates cytolytic effector function in CD8+ T cells. J Immunol 182 , 3380-3389.

Choi, B.D., Suryadevara, C.M., Gedeon, P.C., Herndon, J.E., 2nd, Sanchez-Perez, L., Bigner, D.D., and Sampson, J.H. (2014). Intracerebral delivery of a third generation EGFRvIII-specific chimeric antigen receptor is efficacious against human glioma. Journal of clinical neuroscience : official journal of the Neurosurgical Society of Australasia 21 , 189-190.

Chow, K.K., Naik, S., Kakarla, S., Brawley, V.S., Shaffer, D.R., Yi, Z., Rainusso, N., Wu, M.F., Liu, H., Kew, Y. , et al. (2013). T cells redirected to EphA2 for the 189

immunotherapy of glioblastoma. Molecular therapy : the journal of the American Society of Gene Therapy 21 , 629-637.

Ciceri, F., Bonini, C., Stanghellini, M.T., Bondanza, A., Traversari, C., Salomoni, M., Turchetto, L., Colombi, S., Bernardi, M., Peccatori, J. , et al. (2009). Infusion of suicide- gene-engineered donor lymphocytes after family haploidentical haemopoietic stem-cell transplantation for leukaemia (the TK007 trial): a non-randomised phase I-II study. The Lancet. Oncology 10 , 489-500.

Cieri, N., Camisa, B., Cocchiarella, F., Forcato, M., Oliveira, G., Provasi, E., Bondanza, A., Bordignon, C., Peccatori, J., Ciceri, F. , et al. (2013). IL-7 and IL-15 instruct the generation of human memory stem T cells from naive precursors. Blood 121 , 573-584.

Comi, C., Fleetwood, T., and Dianzani, U. (2012). The role of T cell apoptosis in nervous system autoimmunity. Autoimmun Rev 12 , 150-156.

Connolly, E.C., Freimuth, J., and Akhurst, R.J. (2012). Complexities of TGF-beta targeted cancer therapy. Int J Biol Sci 8, 964-978.

Cooper, G.S., and Stroehla, B.C. (2003). The epidemiology of autoimmune diseases. Autoimmunity Reviews 2, 119-125.

Corazza, N., Muller, S., Brunner, T., Kagi, D., and Mueller, C. (2000). Differential contribution of Fas- and perforin-mediated mechanisms to the cell-mediated cytotoxic activity of naive and in vivo-primed intestinal intraepithelial lymphocytes. J Immunol 164 , 398-403.

Crespo, J., Sun, H., Welling, T.H., Tian, Z., and Zou, W. (2013). T cell anergy, exhaustion, senescence, and stemness in the tumor microenvironment. Curr Opin Immunol 25 , 214-221.

Cruz-Guilloty F, P.M., Djuretic IM, Levanon D, Lotem J, Lichtenheld MG, Groner Y, Rao A. (2009). Runx3 and T-box proteins cooperate to establish the transcriptional program of effector CTLs. J Exp Med 206 , 51-59.

190

Cui, G., Hara, T., Simmons, S., Wagatsuma, K., Abe, A., Miyachi, H., Kitano, S., Ishii, M., Tani-ichi, S., and Ikuta, K. (2014). Characterization of the IL-15 niche in primary and secondary lymphoid organs in vivo. Proc Natl Acad Sci U S A 111 , 1915-1920.

Curiel, T.J., Wei, S., Dong, H., Alvarez, X., Cheng, P., Mottram, P., Krzysiek, R., Knutson, K.L., Daniel, B., Zimmermann, M.C. , et al. (2003). Blockade of B7-H1 improves myeloid dendritic cell-mediated antitumor immunity. Nat Med 9, 562-567.

Curran, M.A., Kim, M., Montalvo, W., Al-Shamkhani, A., and Allison, J.P. (2011). Combination CTLA-4 blockade and 4-1BB activation enhances tumor rejection by increasing T-cell infiltration, proliferation, and cytokine production. PLoS One 6, e19499.

Davis, I.D., Brady, B., Kefford, R.F., Millward, M., Cebon, J., Skrumsager, B.K., Mouritzen, U., Hansen, L.T., Skak, K., Lundsgaard, D. , et al. (2009). Clinical and biological efficacy of recombinant human interleukin-21 in patients with stage IV malignant melanoma without prior treatment: a phase IIa trial. Clin Cancer Res 15 , 2123- 2129.

Davis, I.D., Skrumsager, B.K., Cebon, J., Nicholaou, T., Barlow, J.W., Moller, N.P., Skak, K., Lundsgaard, D., Frederiksen, K.S., Thygesen, P., and McArthur, G.A. (2007). An open-label, two-arm, phase I trial of recombinant human interleukin-21 in patients with metastatic melanoma. Clin Cancer Res 13 , 3630-3636.

Davis, J.E., Smyth, M.J., and Trapani, J.A. (2001). Granzyme A and B-deficient killer lymphocytes are defective in eliciting DNA fragmentation but retain potent in vivo anti- tumor capacity. Eur J Immunol 31 , 39-47.

De Jaeger, K., Seppenwoolde, Y., Kampinga, H.H., Boersma, L.J., Belderbos, J.S., and Lebesque, J.V. (2004). Significance of plasma transforming growth factor-beta levels in radiotherapy for non-small-cell lung cancer. Int J Radiat Oncol Biol Phys 58 , 1378-1387.

De Jaeger K, S.Y., Kampinga HH, Boersma LJ, Belderbos JS, Lebesque JV (2004). Significance of plasma transforming growth factor-beta levels in radiotherapy for non- small-cell lung cancer. Int J Radiat Oncol Biol Phys 58 , 1378-1387. de Verteuil, D., Muratore-Schroeder, T.L., Granados, D.P., Fortier, M.H., Hardy, M.P., Bramoulle, A., Caron, E., Vincent, K., Mader, S., Lemieux, S. , et al. (2010). Deletion of 191

immunoproteasome subunits imprints on the transcriptome and has a broad impact on peptides presented by major histocompatibility complex I molecules. Molecular & cellular proteomics : MCP 9, 2034-2047.

Demeure, C.E., Wolfers, J., Martin-Garcia, N., Gaulard, P., and Triebel, F. (2001). T Lymphocytes infiltrating various tumour types express the MHC class II ligand lymphocyte activation gene-3 (LAG-3): role of LAG-3/MHC class II interactions in cell- cell contacts. European journal of cancer 37 , 1709-1718.

Deng, L., Liang, H., Xu, M., Yang, X., Burnette, B., Arina, A., Li, X.D., Mauceri, H., Beckett, M., Darga, T. , et al. (2014). STING-Dependent Cytosolic DNA Sensing Promotes Radiation-Induced Type I Interferon-Dependent Antitumor Immunity in Immunogenic Tumors. Immunity 41 , 843-852.

Derynck, R., Jarrett, J.A., Chen, E.Y., Eaton, D.H., Bell, J.R., Assoian, R.K., Roberts, A.B., Sporn, M.B., and Goeddel, D.V. (1985). Human transforming growth factor-beta complementary DNA sequence and expression in normal and transformed cells. Nature 316 , 701-705.

Derynck R, J.J., Chen EY, Eaton DH, Bell JR, Assoian RK, Roberts AB, Sporn MB, Goeddel DV (1985). Human transforming growth factor-beta complementary DNA sequence and expression in normal and transformed cells. Nature 316 , 701-705.

Di Stasi, A., Tey, S.K., Dotti, G., Fujita, Y., Kennedy-Nasser, A., Martinez, C., Straathof, K., Liu, E., Durett, A.G., Grilley, B. , et al. (2011). Inducible apoptosis as a safety switch for adoptive cell therapy. N Engl J Med 365 , 1673-1683.

Doubrovina, E.S., Doubrovin, M.M., Vider, E., Sisson, R.B., O'Reilly, R.J., Dupont, B., and Vyas, Y.M. (2003). Evasion from NK cell immunity by MHC class I chain-related molecules expressing colon adenocarcinoma. J Immunol 171 , 6891-6899.

Dubois, S., Mariner, J., Waldmann, T.A., and Tagaya, Y. (2002). IL-15Ralpha recycles and presents IL-15 In trans to neighboring cells. Immunity 17 , 537-547.

Dubois, S., Patel, H.J., Zhang, M., Waldmann, T.A., and Muller, J.R. (2008). Preassociation of IL-15 with IL-15R alpha-IgG1-Fc enhances its activity on proliferation of NK and CD8+/CD44high T cells and its antitumor action. J Immunol 180 , 2099-2106.

192

Dudda, J.C., Salaun, B., Ji, Y., Palmer, D.C., Monnot, G.C., Merck, E., Boudousquie, C., Utzschneider, D.T., Escobar, T.M., Perret, R. , et al. (2013). MicroRNA-155 is required for effector CD8+ T cell responses to virus infection and cancer. Immunity 38 , 742-753.

Dudley, M.E., Wunderlich, J.R., Robbins, P.F., Yang, J.C., Hwu, P., Schwartzentruber, D.J., Topalian, S.L., Sherry, R., Restifo, N.P., Hubicki, A.M. , et al. (2002). Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science 298 , 850-854.

Ebert, M.S., Neilson, J.R., and Sharp, P.A. (2007). MicroRNA sponges: competitive inhibitors of small RNAs in mammalian cells. Nat Methods 4, 721-726.

Ebert, M.S., and Sharp, P.A. (2010). MicroRNA sponges: progress and possibilities. Rna 16 , 2043-2050.

Ebert PJ, J.S., Xie J, Li QJ, Davis MM. (2009). An endogenous positively selecting peptide enhances mature T cell responses and becomes an autoantigen in the absence of microRNA miR-181a. Nat Immunol 10 , 1162-1169.

Elgueta, R., Benson, M.J., de Vries, V.C., Wasiuk, A., Guo, Y., and Noelle, R.J. (2009). Molecular mechanism and function of CD40/CD40L engagement in the immune system. Immunological reviews 229 , 152-172.

Engelhardt, B. (2008). The blood-central nervous system barriers actively control immune cell entry into the central nervous system. Current pharmaceutical design 14 , 1555-1565.

Eshhar, Z., Waks, T., Gross, G., and Schindler, D.G. (1993). Specific activation and targeting of cytotoxic lymphocytes through chimeric single chains consisting of antibody- binding domains and the gamma or zeta subunits of the immunoglobulin and T-cell receptors. Proc Natl Acad Sci U S A 90 , 720-724.

Essand, M., and Loskog, A.S. (2013). Genetically engineered T cells for the treatment of cancer. Journal of internal medicine 273 , 166-181.

193

Fauriat, C., Long, E.O., Ljunggren, H.G., and Bryceson, Y.T. (2010). Regulation of human NK-cell cytokine and chemokine production by target cell recognition. Blood 115 , 2167-2176.

Fernandez, D., and Perl, A. (2009). Metabolic control of T cell activation and death in SLE. Autoimmun Rev 8, 184-189.

Fife, B.T., and Pauken, K.E. (2011). The role of the PD-1 pathway in autoimmunity and peripheral tolerance. Ann N Y Acad Sci 1217 , 45-59.

Filonov, G.S., Piatkevich, K.D., Ting, L.M., Zhang, J., Kim, K., and Verkhusha, V.V. (2011). Bright and stable near-infrared fluorescent protein for in vivo imaging. Nat Biotechnol 29 , 757-761.

Flavell, R.A., Sanjabi, S., Wrzesinski, S.H., and Licona-Limon, P. (2010). The polarization of immune cells in the tumour environment by TGFbeta. Nat Rev Immunol 10 , 554-567.

Fleisher, T.A., and Oliveira, J.B. (2012). Monogenic defects in lymphocyte apoptosis. Curr Opin Allergy Clin Immunol 12 , 609-615.

Fong, L., Carroll, P., Weinberg, V., Chan, S., Lewis, J., Corman, J., Amling, C.L., Stephenson, R.A., Simko, J., Sheikh, N.A. , et al. (2014). Activated lymphocyte recruitment into the tumor microenvironment following preoperative sipuleucel-T for localized prostate cancer. Journal of the National Cancer Institute 106 .

Fourcade, J., Sun, Z., Benallaoua, M., Guillaume, P., Luescher, I.F., Sander, C., Kirkwood, J.M., Kuchroo, V., and Zarour, H.M. (2010). Upregulation of Tim-3 and PD-1 expression is associated with tumor antigen-specific CD8+ T cell dysfunction in melanoma patients. J Exp Med 207 , 2175-2186.

Fourcade J, S.Z., Benallaoua M, Guillaume P, Luescher IF, Sander C, Kirkwood JM, Kuchroo V, Zarour HM (2010). Upregulation of Tim-3 and PD-1 expression is associated with tumor antigen-specific CD8+ T cell dysfunction in melanoma patients. J Exp Med 207 , 2175-2186.

194

Fox, C.J., Hammerman, P.S., and Thompson, C.B. (2005). Fuel feeds function: energy metabolism and the T-cell response. Nat Rev Immunol 5, 844-852.

Frederiksen, K.S., Lundsgaard, D., Freeman, J.A., Hughes, S.D., Holm, T.L., Skrumsager, B.K., Petri, A., Hansen, L.T., McArthur, G.A., Davis, I.D., and Skak, K. (2008). IL-21 induces in vivo immune activation of NK cells and CD8(+) T cells in patients with metastatic melanoma and renal cell carcinoma. Cancer Immunol Immunother 57 , 1439-1449.

Fuchs, E.J., and Matzinger, P. (1992). B cells turn off virgin but not memory T cells. Science 258 , 1156-1159.

Fuertes, M.B., Woo, S.R., Burnett, B., Fu, Y.X., and Gajewski, T.F. (2013). Type I interferon response and innate immune sensing of cancer. Trends Immunol 34 , 67-73.

Gallucci, S., Lolkema, M., and Matzinger, P. (1999). Natural adjuvants: endogenous activators of dendritic cells. Nat Med 5, 1249-1255.

Galon, J., Costes, A., Sanchez-Cabo, F., Kirilovsky, A., Mlecnik, B., Lagorce-Pages, C., Tosolini, M., Camus, M., Berger, A., Wind, P. , et al. (2006). Type, density, and location of immune cells within human colorectal tumors predict clinical outcome. Science 313 , 1960-1964.

Gao, P., Tchernyshyov, I., Chang, T.C., Lee, Y.S., Kita, K., Ochi, T., Zeller, K.I., De Marzo, A.M., Van Eyk, J.E., Mendell, J.T., and Dang, C.V. (2009). c-Myc suppression of miR-23a/b enhances mitochondrial glutaminase expression and glutamine metabolism. Nature 458 , 762-765.

Garrido, F., Algarra, I., and Garcia-Lora, A.M. (2010). The escape of cancer from T lymphocytes: immunoselection of MHC class I loss variants harboring structural- irreversible "hard" lesions. Cancer Immunol Immunother 59 , 1601-1606.

Gattinoni, L., Lugli, E., Ji, Y., Pos, Z., Paulos, C.M., Quigley, M.F., Almeida, J.R., Gostick, E., Yu, Z., Carpenito, C. , et al. (2011). A human memory T cell subset with stem cell-like properties. Nat Med 17 , 1290-1297.

195

Gatzka, M., and Walsh, C.M. (2007). Apoptotic signal transduction and T cell tolerance. Autoimmunity 40 , 442-452.

Genestier, L., Kasibhatla, S., Brunner, T., and Green, D.R. (1999). Transforming growth factor beta1 inhibits Fas ligand expression and subsequent activation-induced cell death in T cells via downregulation of c-Myc. J Exp Med 189 , 231-239.

Gerriets, V.A., and Rathmell, J.C. (2012). Metabolic pathways in T cell fate and function. Trends Immunol 33 , 168-173.

Getts, D.R., Shankar, S., Chastain, E.M., Martin, A., Getts, M.T., Wood, K., and Miller, S.D. (2011). Current landscape for T-cell targeting in autoimmunity and transplantation. Immunotherapy 3, 853-870.

Gilboa, E. (2004). The promise of cancer vaccines. Nat Rev Cancer 4, 401-411.

Girling, A., Bartkova, J., Burchell, J., Gendler, S., Gillett, C., and Taylor-Papadimitriou, J. (1989). A core protein epitope of the polymorphic epithelial mucin detected by the monoclonal antibody SM-3 is selectively exposed in a range of primary carcinomas. Int J Cancer 43 , 1072-1076.

Goodnow, C.C., Sprent, J., Fazekas de St Groth, B., and Vinuesa, C.G. (2005). Cellular and genetic mechanisms of self tolerance and autoimmunity. Nature 435 , 590-597.

Gorelik, L., Constant, S., and Flavell, R.A. (2002). Mechanism of transforming growth factor beta-induced inhibition of T helper type 1 differentiation. J Exp Med 195 , 1499- 1505.

Gracias, D.T., Stelekati, E., Hope, J.L., Boesteanu, A.C., Doering, T.A., Norton, J., Mueller, Y.M., Fraietta, J.A., Wherry, E.J., Turner, M., and Katsikis, P.D. (2013). The microRNA miR-155 controls CD8(+) T cell responses by regulating interferon signaling. Nat Immunol 14 , 593-602.

Graef, P., Buchholz, V.R., Stemberger, C., Flossdorf, M., Henkel, L., Schiemann, M., Drexler, I., Hofer, T., Riddell, S.R., and Busch, D.H. (2014). Serial transfer of single- cell-derived immunocompetence reveals stemness of CD8(+) central memory T cells. Immunity 41 , 116-126. 196

Grimaldi, M., Denizot, M., Espert, L., Robert-Hebmann, V., and Biard-Piechaczyk, M. (2005). Mitochondria-dependent apoptosis in T-cell homeostasis. Current opinion in investigational drugs 6, 1095-1102.

Grosso, J.F., Kelleher, C.C., Harris, T.J., Maris, C.H., Hipkiss, E.L., De Marzo, A., Anders, R., Netto, G., Getnet, D., Bruno, T.C. , et al. (2007). LAG-3 regulates CD8+ T cell accumulation and effector function in murine self- and tumor-tolerance systems. J Clin Invest 117 , 3383-3392.

Gu, A.D., Wang, Y., Lin, L., Zhang, S.S., and Wan, Y.Y. (2012). Requirements of transcription factor Smad-dependent and -independent TGF-beta signaling to control discrete T-cell functions. Proc Natl Acad Sci U S A 109 , 905-910.

Guerra, N., Tan, Y.X., Joncker, N.T., Choy, A., Gallardo, F., Xiong, N., Knoblaugh, S., Cado, D., Greenberg, N.M., and Raulet, D.H. (2008). NKG2D-deficient mice are defective in tumor surveillance in models of spontaneous malignancy. Immunity 28 , 571- 580.

Gupta, S., and Gollapudi, S. (2007). Susceptibility of naive and subsets of memory T cells to apoptosis via multiple signaling pathways. Autoimmun Rev 6, 476-481.

Hamanishi, J., Mandai, M., Iwasaki, M., Okazaki, T., Tanaka, Y., Yamaguchi, K., Higuchi, T., Yagi, H., Takakura, K., Minato, N. , et al. (2007). Programmed cell death 1 ligand 1 and tumor-infiltrating CD8+ T lymphocytes are prognostic factors of human ovarian cancer. Proc Natl Acad Sci U S A 104 , 3360-3365.

Hamid, O., Robert, C., Daud, A., Hodi, F.S., Hwu, W.J., Kefford, R., Wolchok, J.D., Hersey, P., Joseph, R.W., Weber, J.S. , et al. (2013). Safety and tumor responses with lambrolizumab (anti-PD-1) in melanoma. N Engl J Med 369 , 134-144.

Han, S., Zhang, C., Li, Q., Dong, J., Liu, Y., Huang, Y., Jiang, T., and Wu, A. (2014). Tumour-infiltrating CD4(+) and CD8(+) lymphocytes as predictors of clinical outcome in glioma. Br J Cancer 110 , 2560-2568.

Hawiger, D., Inaba, K., Dorsett, Y., Guo, M., Mahnke, K., Rivera, M., Ravetch, J.V., Steinman, R.M., and Nussenzweig, M.C. (2001). Dendritic cells induce peripheral T cell unresponsiveness under steady state conditions in vivo. J Exp Med 194 , 769-779.

197

Heimberger, A.B., Hlatky, R., Suki, D., Yang, D., Weinberg, J., Gilbert, M., Sawaya, R., and Aldape, K. (2005). Prognostic effect of epidermal growth factor receptor and EGFRvIII in glioblastoma multiforme patients. Clinical cancer research : an official journal of the American Association for Cancer Research 11 , 1462-1466.

Hernandez-Chacon, J.A., Li, Y., Wu, R.C., Bernatchez, C., Wang, Y., Weber, J.S., Hwu, P., and Radvanyi, L.G. (2011). Costimulation through the CD137/4-1BB pathway protects human melanoma tumor-infiltrating lymphocytes from activation-induced cell death and enhances antitumor effector function. J Immunother 34 , 236-250.

Higano, C.S., Schellhammer, P.F., Small, E.J., Burch, P.A., Nemunaitis, J., Yuh, L., Provost, N., and Frohlich, M.W. (2009). Integrated data from 2 randomized, double- blind, placebo-controlled, phase 3 trials of active cellular immunotherapy with sipuleucel-T in advanced prostate cancer. Cancer 115 , 3670-3679.

Hinrichs, C.S., Spolski, R., Paulos, C.M., Gattinoni, L., Kerstann, K.W., Palmer, D.C., Klebanoff, C.A., Rosenberg, S.A., Leonard, W.J., and Restifo, N.P. (2008). IL-2 and IL- 21 confer opposing differentiation programs to CD8+ T cells for adoptive immunotherapy. Blood 111 , 5326-5333.

Hivroz, C., Chemin, K., Tourret, M., and Bohineust, A. (2012). Crosstalk between T lymphocytes and dendritic cells. Critical reviews in immunology 32 , 139-155.

Ho, L.J., and Lai, J.H. (2004). Chinese herbs as immunomodulators and potential disease- modifying antirheumatic drugs in autoimmune disorders. Curr Drug Metab 5, 181-192.

Hsu, C., Abad, J.D., and Morgan, R.A. (2013). Characterization of human T lymphocytes engineered to express interleukin-15 and herpes simplex virus-thymidine kinase. J Surg Res 184 , 282-289.

Hu, X., Chakravarty, S.D., and Ivashkiv, L.B. (2008). Regulation of interferon and Toll- like receptor signaling during macrophage activation by opposing feedforward and feedback inhibition mechanisms. Immunological reviews 226 , 41-56.

Huang, C.T., Workman, C.J., Flies, D., Pan, X., Marson, A.L., Zhou, G., Hipkiss, E.L., Ravi, S., Kowalski, J., Levitsky, H.I. , et al. (2004). Role of LAG-3 in regulatory T cells. Immunity 21 , 503-513.

198

Huard, B., Gaulard, P., Faure, F., Hercend, T., and Triebel, F. (1994). Cellular expression and tissue distribution of the human LAG-3-encoded protein, an MHC class II ligand. Immunogenetics 39 , 213-217.

Huard, B., Prigent, P., Tournier, M., Bruniquel, D., and Triebel, F. (1995). CD4/major histocompatibility complex class II interaction analyzed with CD4- and lymphocyte activation gene-3 (LAG-3)-Ig fusion proteins. Eur J Immunol 25 , 2718-2721.

Huber, E.M., Basler, M., Schwab, R., Heinemeyer, W., Kirk, C.J., Groettrup, M., and Groll, M. (2012). Immuno- and constitutive proteasome crystal structures reveal differences in substrate and inhibitor specificity. Cell 148 , 727-738.

Humphrey, P.A., Wong, A.J., Vogelstein, B., Zalutsky, M.R., Fuller, G.N., Archer, G.E., Friedman, H.S., Kwatra, M.M., Bigner, S.H., and Bigner, D.D. (1990). Anti-synthetic peptide antibody reacting at the fusion junction of deletion-mutant epidermal growth factor receptors in human glioblastoma. Proc Natl Acad Sci U S A 87 , 4207-4211.

Iliopoulou, E.G., Kountourakis, P., Karamouzis, M.V., Doufexis, D., Ardavanis, A., Baxevanis, C.N., Rigatos, G., Papamichail, M., and Perez, S.A. (2010). A phase I trial of adoptive transfer of allogeneic natural killer cells in patients with advanced non-small cell lung cancer. Cancer Immunol Immunother 59 , 1781-1789.

Inda, M.M., Bonavia, R., Mukasa, A., Narita, Y., Sah, D.W., Vandenberg, S., Brennan, C., Johns, T.G., Bachoo, R., Hadwiger, P. , et al. (2010). Tumor heterogeneity is an active process maintained by a mutant EGFR-induced cytokine circuit in glioblastoma. Genes & development 24 , 1731-1745.

Intlekofer AM, B.A., Takemoto N, Gordon SM, Dejong CS, Shin H, Hunter CA, Wherry EJ, Lindsten T, Reiner SL. (2008). Anomalous type 17 response to viral infection by CD8+ T cells lacking T-bet and eomesodermin. Science 321 , 408-411.

Intlekofer, A.M., Takemoto, N., Kao, C., Banerjee, A., Schambach, F., Northrop, J.K., Shen, H., Wherry, E.J., and Reiner, S.L. (2007). Requirement for T-bet in the aberrant differentiation of unhelped memory CD8+ T cells. J Exp Med 204 , 2015-2021.

Intlekofer, A.M., Takemoto, N., Wherry, E.J., Longworth, S.A., Northrup, J.T., Palanivel, V.R., Mullen, A.C., Gasink, C.R., Kaech, S.M., Miller, J.D. , et al. (2005). Effector and

199

memory CD8+ T cell fate coupled by T-bet and eomesodermin. Nat Immunol 6, 1236- 1244.

Ito, M., Minamiya, Y., Kawai, H., Saito, S., Saito, H., Nakagawa, T., Imai, K., Hirokawa, M., and Ogawa, J. (2006). Tumor-derived TGFbeta-1 induces dendritic cell apoptosis in the sentinel lymph node. J Immunol 176 , 5637-5643.

Iwamura K, K.T., Miyahara Y, Naota H, Mineno J, Ikeda H, Shiku H (2012). siRNA- mediated silencing of PD-1 ligands enhances tumor-specific human T-cell effector functions. Gene Ther 19 , 959-966.

Jackson, C., Ruzevick, J., Phallen, J., Belcaid, Z., and Lim, M. (2011). Challenges in immunotherapy presented by the glioblastoma multiforme microenvironment. Clin Dev Immunol 2011 , 732413.

Ji, Y., Pos, Z., Rao, M., Klebanoff, C.A., Yu, Z., Sukumar, M., Reger, R.N., Palmer, D.C., Borman, Z.A., Muranski, P. , et al. (2011). Repression of the DNA-binding inhibitor Id3 by Blimp-1 limits the formation of memory CD8+ T cells. Nat Immunol 12 , 1230- 1237.

Jiang, S., Li, C., Olive, V., Lykken, E., Feng, F., Sevilla, J., Wan, Y., He, L., and Li, Q.J. (2011). Molecular dissection of the miR-17-92 cluster's critical dual roles in promoting Th1 responses and preventing inducible Treg differentiation. Blood 118 , 5487-5497.

Jiang, X., Clark, R.A., Liu, L., Wagers, A.J., Fuhlbrigge, R.C., and Kupper, T.S. (2012). Skin infection generates non-migratory memory CD8+ T(RM) cells providing global skin immunity. Nature 483 , 227-231.

Jing, W., Gershan, J.A., Weber, J., Tlomak, D., McOlash, L., Sabatos-Peyton, C., and Johnson, B.D. (2015). Combined immune checkpoint protein blockade and low dose whole body irradiation as immunotherapy for myeloma. Journal for immunotherapy of cancer 3, 2.

Johnson, S.M., Grosshans, H., Shingara, J., Byrom, M., Jarvis, R., Cheng, A., Labourier, E., Reinert, K.L., Brown, D., and Slack, F.J. (2005). RAS is regulated by the let-7 microRNA family. Cell 120 , 635-647.

200

Joshi, N.S., Cui, W., Chandele, A., Lee, H.K., Urso, D.R., Hagman, J., Gapin, L., and Kaech, S.M. (2007). Inflammation directs memory precursor and short-lived effector CD8(+) T cell fates via the graded expression of T-bet transcription factor. Immunity 27 , 281-295.

Juedes, A.E., Rodrigo, E., Togher, L., Glimcher, L.H., and von Herrath, M.G. (2004). T- bet controls autoaggressive CD8 lymphocyte responses in type 1 diabetes. J Exp Med 199 , 1153-1162.

Kaech, S.M., Hemby, S., Kersh, E., and Ahmed, R. (2002). Molecular and functional profiling of memory CD8 T cell differentiation. Cell 111 , 837-851.

Kallies A, X.A., Belz GT, Nutt SL. (2009). Blimp-1 transcription factor is required for the differentiation of effector CD8(+) T cells and memory responses. Immunity. 31 , 283- 295.

Kallies, A., Xin, A., Belz, G.T., and Nutt, S.L. (2009). Blimp-1 transcription factor is required for the differentiation of effector CD8(+) T cells and memory responses. Immunity 31 , 283-295.

Kalos, M., Levine, B.L., Porter, D.L., Katz, S., Grupp, S.A., Bagg, A., and June, C.H. (2011). T cells with chimeric antigen receptors have potent antitumor effects and can establish memory in patients with advanced leukemia. Sci Transl Med 3, 95ra73.

Kaneko, S., Mastaglio, S., Bondanza, A., Ponzoni, M., Sanvito, F., Aldrighetti, L., Radrizzani, M., La Seta-Catamancio, S., Provasi, E., Mondino, A. , et al. (2009). IL-7 and IL-15 allow the generation of suicide gene-modified alloreactive self-renewing central memory human T lymphocytes. Blood 113 , 1006-1015.

Kanellopoulou, C., Muljo, S.A., Kung, A.L., Ganesan, S., Drapkin, R., Jenuwein, T., Livingston, D.M., and Rajewsky, K. (2005). Dicer-deficient mouse embryonic stem cells are defective in differentiation and centromeric silencing. Genes Dev 19 , 489-501.

Kantoff, P.W., Higano, C.S., Shore, N.D., Berger, E.R., Small, E.J., Penson, D.F., Redfern, C.H., Ferrari, A.C., Dreicer, R., Sims, R.B. , et al. (2010). Sipuleucel-T immunotherapy for castration-resistant prostate cancer. N Engl J Med 363 , 411-422.

201

Kaplan, D.H., Shankaran, V., Dighe, A.S., Stockert, E., Aguet, M., Old, L.J., and Schreiber, R.D. (1998). Demonstration of an interferon gamma-dependent tumor surveillance system in immunocompetent mice. Proc Natl Acad Sci U S A 95 , 7556- 7561.

Kassahn, D., Nachbur, U., Conus, S., Micheau, O., Schneider, P., Simon, H.U., and Brunner, T. (2009). Distinct requirements for activation-induced cell surface expression of preformed Fas/CD95 ligand and cytolytic granule markers in T cells. Cell Death Differ 16 , 115-124.

Kastenmuller, K., Wille-Reece, U., Lindsay, R.W., Trager, L.R., Darrah, P.A., Flynn, B.J., Becker, M.R., Udey, M.C., Clausen, B.E., Igyarto, B.Z. , et al. (2011). Protective T cell immunity in mice following protein-TLR7/8 agonist-conjugate immunization requires aggregation, type I IFN, and multiple DC subsets. J Clin Invest 121 , 1782-1796.

Kawakami, Y., Eliyahu, S., Delgado, C.H., Robbins, P.F., Rivoltini, L., Topalian, S.L., Miki, T., and Rosenberg, S.A. (1994). Cloning of the gene coding for a shared human melanoma antigen recognized by autologous T cells infiltrating into tumor. Proc Natl Acad Sci U S A 91 , 3515-3519.

Keir, M.E., Liang, S.C., Guleria, I., Latchman, Y.E., Qipo, A., Albacker, L.A., Koulmanda, M., Freeman, G.J., Sayegh, M.H., and Sharpe, A.H. (2006). Tissue expression of PD-L1 mediates peripheral T cell tolerance. J Exp Med 203 , 883-895.

Kerkar, S.P., Sanchez-Perez, L., Yang, S., Borman, Z.A., Muranski, P., Ji, Y., Chinnasamy, D., Kaiser, A.D., Hinrichs, C.S., Klebanoff, C.A. , et al. (2011). Genetic engineering of murine CD8+ and CD4+ T cells for preclinical adoptive immunotherapy studies. J Immunother 34 , 343-352.

Kershaw, M.H., Westwood, J.A., Parker, L.L., Wang, G., Eshhar, Z., Mavroukakis, S.A., White, D.E., Wunderlich, J.R., Canevari, S., Rogers-Freezer, L. , et al. (2006). A phase I study on adoptive immunotherapy using gene-modified T cells for ovarian cancer. Clin Cancer Res 12 , 6106-6115.

Kershaw, M.H., Westwood, J.A., Slaney, C.Y., and Darcy, P.K. (2014). Clinical application of genetically modified T cells in cancer therapy. Clinical & translational immunology 3, e16.

202

Khan, A.A., Penny, L.A., Yuzefpolskiy, Y., Sarkar, S., and Kalia, V. (2013). MicroRNA- 17~92 regulates effector and memory CD8 T-cell fates by modulating proliferation in response to infections. Blood 121 , 4473-4483.

Kikuchi, T., Mineta, M., Ohtaka, J., Matsumoto, N., and Katoh, T. (2011). Enantioselective Total Synthesis of (-)-Subglutinols A and B: Potential Immunosuppressive Agents Isolated from a Microorganism. European Journal of Organic Chemistry 2011 , 5020-5030, S5020/5021-S5020/5029.

Kilinc, M.O., Gu, T., Harden, J.L., Virtuoso, L.P., and Egilmez, N.K. (2009). Central role of tumor-associated CD8+ T effector/memory cells in restoring systemic antitumor immunity. J Immunol 182 , 4217-4225.

Kim, H., Baker, J.B., Lee, S.U., Park, Y., Bolduc, K.L., Park, H.B., Dickens, M.G., Lee, D.S., Kim, Y., Kim, S.H., and Hong, J. (2009). Stereoselective synthesis and osteogenic activity of subglutinols A and B. J Am Chem Soc 131 , 3192-3194.

Kim, H., Baker, J.B., Park, Y., Park, H.B., DeArmond, P.D., Kim, S.H., Fitzgerald, M.C., Lee, D.S., and Hong, J. (2010). Total synthesis, assignment of the absolute stereochemistry, and structure-activity relationship studies of subglutinols A and B. Chemistry, an Asian journal 5, 1902-1910.

Kim, Y.H., Jung, T.Y., Jung, S., Jang, W.Y., Moon, K.S., Kim, I.Y., Lee, M.C., and Lee, J.J. (2012). Tumour-infiltrating T-cell subpopulations in glioblastomas. British journal of neurosurgery 26 , 21-27.

Kirkwood, J.M., Lorigan, P., Hersey, P., Hauschild, A., Robert, C., McDermott, D., Marshall, M.A., Gomez-Navarro, J., Liang, J.Q., and Bulanhagui, C.A. (2010). Phase II trial of tremelimumab (CP-675,206) in patients with advanced refractory or relapsed melanoma. Clin Cancer Res 16 , 1042-1048.

Klebanoff, C.A., Finkelstein, S.E., Surman, D.R., Lichtman, M.K., Gattinoni, L., Theoret, M.R., Grewal, N., Spiess, P.J., Antony, P.A., Palmer, D.C. , et al. (2004). IL-15 enhances the in vivo antitumor activity of tumor-reactive CD8+ T cells. Proc Natl Acad Sci U S A 101 , 1969-1974.

Klebanoff, C.A., Gattinoni, L., Palmer, D.C., Muranski, P., Ji, Y., Hinrichs, C.S., Borman, Z.A., Kerkar, S.P., Scott, C.D., Finkelstein, S.E. , et al. (2011). Determinants of 203

successful CD8+ T-cell adoptive immunotherapy for large established tumors in mice. Clin Cancer Res 17 , 5343-5352.

Kloetzel, P.M. (2001). Antigen processing by the proteasome. Nature reviews. Molecular cell biology 2, 179-187.

Knorr, D.A., Bachanova, V., Verneris, M.R., and Miller, J.S. (2014). Clinical utility of natural killer cells in cancer therapy and transplantation. Seminars in immunology 26 , 161-172.

Kobayashi, K., Kaneda, K., and Kasama, T. (2001). Immunopathogenesis of delayed- type hypersensitivity. Microsc Res Tech 53 , 241-245.

Kochenderfer, J.N., Dudley, M.E., Feldman, S.A., Wilson, W.H., Spaner, D.E., Maric, I., Stetler-Stevenson, M., Phan, G.Q., Hughes, M.S., Sherry, R.M. , et al. (2012). B-cell depletion and remissions of malignancy along with cytokine-associated toxicity in a clinical trial of anti-CD19 chimeric-antigen-receptor-transduced T cells. Blood 119 , 2709-2720.

Kohler, S., Wagner, U., Pierer, M., Kimmig, S., Oppmann, B., Mowes, B., Julke, K., Romagnani, C., and Thiel, A. (2005). Post-thymic in vivo proliferation of naive CD4+ T cells constrains the TCR repertoire in healthy human adults. Eur J Immunol 35 , 1987- 1994.

Kong F, J.R., Huang DH, Clough RW, Anscher MS (1999). Plasma transforming growth factor-beta1 level before radiotherapy correlates with long term outcome of patients with lung carcinoma. Cancer 86 , 1712-1719.

Kong, S., Sengupta, S., Tyler, B., Bais, A.J., Ma, Q., Doucette, S., Zhou, J., Sahin, A., Carter, B.S., Brem, H. , et al. (2012). Suppression of human glioma xenografts with second-generation IL13R-specific chimeric antigen receptor-modified T cells. Clin Cancer Res 18 , 5949-5960.

Koralov SB, M.S., Galler GR, Krek A, Chakraborty T, Kanellopoulou C, Jensen K, Cobb BS, Merkenschlager M, Rajewsky N, Rajewsky K. (2008 ). Dicer ablation affects antibody diversity and cell survival in the B lymphocyte lineage. Cell 132 , 860-874.

204

Kota J, C.R., O'Donnell KA, Wentzel EA, Montgomery CL, Hwang HW, Chang TC, Vivekanandan P, Torbenson M, Clark KR, Mendell JR, Mendell JT. (2009). Therapeutic microRNA delivery suppresses tumorigenesis in a murine liver cancer model. Cell 137 , 1005-1017.

Krasagakis K, T.D., Farthmann B, Eberle J, Mansmann U, Orfanos CE (1998). Elevated plasma levels of transforming growth factor (TGF)-beta1 and TGF-beta2 in patients with disseminated malignant melanoma. Br J Cancer 77 , 1492-1494.

Krasagakis, K., Tholke, D., Farthmann, B., Eberle, J., Mansmann, U., and Orfanos, C.E. (1998). Elevated plasma levels of transforming growth factor (TGF)-beta1 and TGF- beta2 in patients with disseminated malignant melanoma. Br J Cancer 77 , 1492-1494.

Krutzfeldt, J., Rajewsky, N., Braich, R., Rajeev, K.G., Tuschl, T., Manoharan, M., and Stoffel, M. (2005). Silencing of microRNAs in vivo with 'antagomirs'. Nature 438 , 685- 689.

Krützfeldt J, R.N., Braich R, Rajeev KG, Tuschl T, Manoharan M, Stoffel M. (2005). Silencing of microRNAs in vivo with 'antagomirs'. Nature 438 , 685-689.

Kwissa, M., Nakaya, H.I., Oluoch, H., and Pulendran, B. (2012). Distinct TLR adjuvants differentially stimulate systemic and local innate immune responses in nonhuman primates. Blood 119 , 2044-2055.

Lamers, C.H., Sleijfer, S., Vulto, A.G., Kruit, W.H., Kliffen, M., Debets, R., Gratama, J.W., Stoter, G., and Oosterwijk, E. (2006). Treatment of metastatic renal cell carcinoma with autologous T-lymphocytes genetically retargeted against carbonic anhydrase IX: first clinical experience. J Clin Oncol 24 , e20-22.

Lammering, G., Hewit, T.H., Holmes, M., Valerie, K., Hawkins, W., Lin, P.-S., Mikkelsen, R.B., and Schmidt-Ullrich, R.K. (2004). Inhibition of the type III epidermal growth factor receptor variant mutant receptor by dominant-negative EGFR-CD533 enhances malignant glioma cell radiosensitivity. Clinical cancer research : an official journal of the American Association for Cancer Research 10 , 6732-6743.

Lampson, L.A. (2011). Monoclonal antibodies in neuro-oncology: Getting past the blood- brain barrier. mAbs 3, 153-160.

205

Lan, R.Y., Selmi, C., and Gershwin, M.E. (2008). The regulatory, inflammatory, and T cell programming roles of interleukin-2 (IL-2). J Autoimmun 31 , 7-12.

Lee, J.C., Lobkovsky, E., Pliam, N.B., Strobel, G., and Clardy, J. (1995). Subglutinols A and B: Immunosuppressive compounds from the endophytic fungus Fusarium subglutinans. Journal of Organic Chemistry 60 , 7076-7077.

Lee, S.H., Shin, M.S., Park, W.S., Kim, S.Y., Kim, H.S., Han, J.Y., Park, G.S., Dong, S.M., Pi, J.H., Kim, C.S. , et al. (1999). Alterations of Fas (Apo-1/CD95) gene in non- small cell lung cancer. Oncogene 18 , 3754-3760.

Lee, W.G., Kim, W.S., Park, S.G., Kim, H., Hong, J., Ko, H., and Kim, Y.C. (2012). Immunosuppressive effects of subglutinol derivatives. ChemMedChem 7, 218-222.

Leithauser, F., Dhein, J., Mechtersheimer, G., Koretz, K., Bruderlein, S., Henne, C., Schmidt, A., Debatin, K.M., Krammer, P.H., and Moller, P. (1993). Constitutive and induced expression of APO-1, a new member of the nerve growth factor/tumor necrosis factor receptor superfamily, in normal and neoplastic cells. Laboratory investigation; a journal of technical methods and pathology 69 , 415-429.

Leung, S., Liu, X., Fang, L., Chen, X., Guo, T., and Zhang, J. (2010). The cytokine milieu in the interplay of pathogenic Th1/Th17 cells and regulatory T cells in autoimmune disease. Cell Mol Immunol 7, 182-189.

Li, Q., Rao, R.R., Araki, K., Pollizzi, K., Odunsi, K., Powell, J.D., and Shrikant, P.A. (2011). A central role for mTOR kinase in homeostatic proliferation induced CD8+ T cell memory and tumor immunity. Immunity 34 , 541-553.

Li, Q.J., Chau, J., Ebert, P.J., Sylvester, G., Min, H., Liu, G., Braich, R., Manoharan, M., Soutschek, J., Skare, P. , et al. (2007). miR-181a Is an Intrinsic Modulator of T Cell Sensitivity and Selection. Cell 129 , 147-161.

Li QJ, C.J., Ebert PJ, Sylvester G, Min H, Liu G, Braich R, Manoharan M, Soutschek J, Skare P, Klein LO, Davis MM, Chen CZ. (2007). miR-181a is an intrinsic modulator of T cell sensitivity and selection. Cell 129 .

206

Linnemann, C., van Buuren, M.M., Bies, L., Verdegaal, E.M., Schotte, R., Calis, J.J., Behjati, S., Velds, A., Hilkmann, H., Atmioui, D.E., et al. (2015). High-throughput epitope discovery reveals frequent recognition of neo-antigens by CD4+ T cells in human melanoma. Nat Med 21 , 81-85.

Linsley, P.S., Greene, J.L., Brady, W., Bajorath, J., Ledbetter, J.A., and Peach, R. (1994). Human B7-1 (CD80) and B7-2 (CD86) bind with similar avidities but distinct kinetics to CD28 and CTLA-4 receptors. Immunity 1, 793-801.

Lipowska-Bhalla, G., Gilham, D.E., Hawkins, R.E., and Rothwell, D.G. (2012). Targeted immunotherapy of cancer with CAR T cells: achievements and challenges. Cancer Immunol Immunother 61 , 953-962.

Listopad, J.J., Kammertoens, T., Anders, K., Silkenstedt, B., Willimsky, G., Schmidt, K., Kuehl, A.A., Loddenkemper, C., and Blankenstein, T. (2013). Fas expression by tumor stroma is required for cancer eradication. Proc Natl Acad Sci U S A 110 , 2276-2281.

Liu, R.R., and Murphy, T.H. (2009). Reversible cyclosporin A-sensitive mitochondrial depolarization occurs within minutes of stroke onset in mouse somatosensory cortex in vivo: a two-photon imaging study. J Biol Chem 284 , 36109-36117.

Liu, V.C., Wong, L.Y., Jang, T., Shah, A.H., Park, I., Yang, X., Zhang, Q., Lonning, S., Teicher, B.A., and Lee, C. (2007). Tumor evasion of the immune system by converting CD4+CD25- T cells into CD4+CD25+ T regulatory cells: role of tumor-derived TGF- beta. J Immunol 178 , 2883-2892.

Lockman, P.R., Mittapalli, R.K., Taskar, K.S., Rudraraju, V., Gril, B., Bohn, K.A., Adkins, C.E., Roberts, A., Thorsheim, H.R., Gaasch, J.A. , et al. (2010). Heterogeneous blood-tumor barrier permeability determines drug efficacy in experimental brain metastases of breast cancer. Clin Cancer Res 16 , 5664-5678.

Lonning, S., Mannick, J., and McPherson, J.M. (2011). Antibody targeting of TGF-beta in cancer patients. Curr Pharm Biotechnol 12 , 2176-2189.

Lopez, J.A., Brennan, A.J., Whisstock, J.C., Voskoboinik, I., and Trapani, J.A. (2012). Protecting a serial killer: pathways for perforin trafficking and self-defence ensure sequential target cell death. Trends Immunol 33 , 406-412.

207

Lu, J., Getz, G., Miska, E.A., Alvarez-Saavedra, E., Lamb, J., Peck, D., Sweet-Cordero, A., Ebert, B.L., Mak, R.H., Ferrando, A.A. , et al. (2005). MicroRNA expression profiles classify human cancers. Nature 435 , 834-838.

Lu, P., Youngblood, B.A., Austin, J.W., Mohammed, A.U., Butler, R., Ahmed, R., and Boss, J.M. (2014). Blimp-1 represses CD8 T cell expression of PD-1 using a feedforward transcriptional circuit during acute viral infection. J Exp Med 211 , 515-527.

Macon-Lemaitre, L., and Triebel, F. (2005). The negative regulatory function of the lymphocyte-activation gene-3 co-receptor (CD223) on human T cells. Immunology 115 , 170-178.

Maher, J., Brentjens, R.J., Gunset, G., Riviere, I., and Sadelain, M. (2002). Human T- lymphocyte cytotoxicity and proliferation directed by a single chimeric TCRzeta /CD28 receptor. Nat Biotechnol 20 , 70-75.

Mandal, M., Bandyopadhyay, D., Goepfert, T.M., and Kumar, R. (1998). Interferon- induces expression of cyclin-dependent kinase-inhibitors p21WAF1 and p27Kip1 that prevent activation of cyclin-dependent kinase by CDK-activating kinase (CAK). Oncogene 16 , 217-225.

Markley, J.C., and Sadelain, M. (2010). IL-7 and IL-21 are superior to IL-2 and IL-15 in promoting human T cell-mediated rejection of systemic lymphoma in immunodeficient mice. Blood 115 , 3508-3519.

Martins, G.A., Cimmino, L., Liao, J., Magnusdottir, E., and Calame, K. (2008). Blimp-1 directly represses Il2 and the Il2 activator Fos, attenuating T cell proliferation and survival. J Exp Med 205 , 1959-1965.

Masopust, D., Vezys, V., Marzo, A.L., and Lefrancois, L. (2001). Preferential localization of effector memory cells in nonlymphoid tissue. Science 291 , 2413-2417.

Matsuzaki, J., Gnjatic, S., Mhawech-Fauceglia, P., Beck, A., Miller, A., Tsuji, T., Eppolito, C., Qian, F., Lele, S., Shrikant, P. , et al. (2010). Tumor-infiltrating NY-ESO-1- specific CD8+ T cells are negatively regulated by LAG-3 and PD-1 in human ovarian cancer. Proc Natl Acad Sci U S A 107 , 7875-7880.

208

Mattei, F., Schiavoni, G., Belardelli, F., and Tough, D.F. (2001). IL-15 is expressed by dendritic cells in response to type I IFN, double-stranded RNA, or lipopolysaccharide and promotes dendritic cell activation. J Immunol 167 , 1179-1187.

Maus, M.V., Grupp, S.A., Porter, D.L., and June, C.H. (2014). Antibody-modified T cells: CARs take the front seat for hematologic malignancies. Blood 123 , 2625-2635.

McKarns, S.C., Schwartz, R.H., and Kaminski, N.E. (2004). Smad3 is essential for TGF- beta 1 to suppress IL-2 production and TCR-induced proliferation, but not IL-2-induced proliferation. J Immunol 172 , 4275-4284.

Medema, J.P., de Jong, J., Peltenburg, L.T., Verdegaal, E.M., Gorter, A., Bres, S.A., Franken, K.L., Hahne, M., Albar, J.P., Melief, C.J., and Offringa, R. (2001). Blockade of the granzyme B/perforin pathway through overexpression of the serine protease inhibitor PI-9/SPI-6 constitutes a mechanism for immune escape by tumors. Proc Natl Acad Sci U S A 98 , 11515-11520.

Miller, J.D., van der Most, R.G., Akondy, R.S., Glidewell, J.T., Albott, S., Masopust, D., Murali-Krishna, K., Mahar, P.L., Edupuganti, S., Lalor, S. , et al. (2008). Human effector and memory CD8+ T cell responses to smallpox and yellow fever vaccines. Immunity 28 , 710-722.

Miller, J.S., Soignier, Y., Panoskaltsis-Mortari, A., McNearney, S.A., Yun, G.H., Fautsch, S.K., McKenna, D., Le, C., Defor, T.E., Burns, L.J. , et al. (2005). Successful adoptive transfer and in vivo expansion of human haploidentical NK cells in patients with cancer. Blood 105 , 3051-3057.

Mlecnik, B., Tosolini, M., Kirilovsky, A., Berger, A., Bindea, G., Meatchi, T., Bruneval, P., Trajanoski, Z., Fridman, W.H., Pages, F., and Galon, J. (2011). Histopathologic-based prognostic factors of colorectal cancers are associated with the state of the local immune reaction. J Clin Oncol 29 , 610-618.

Montgomery, R.B., Guzman, J., O'Rourke, D.M., and Stahl, W.L. (2000). Expression of oncogenic epidermal growth factor receptor family kinases induces paclitaxel resistance and alters beta-tubulin isotype expression. The Journal of biological chemistry 275 , 17358-17363.

209

Moon, E.K., Wang, L.C., Dolfi, D.V., Wilson, C.B., Ranganathan, R., Sun, J., Kapoor, V., Scholler, J., Pure, E., Milone, M.C. , et al. (2014). Multifactorial T-cell hypofunction that is reversible can limit the efficacy of chimeric antigen receptor-transduced human T cells in solid tumors. Clin Cancer Res 20 , 4262-4273.

Morgan, R.A., Dudley, M.E., Wunderlich, J.R., Hughes, M.S., Yang, J.C., Sherry, R.M., Royal, R.E., Topalian, S.L., Kammula, U.S., Restifo, N.P. , et al. (2006). Cancer regression in patients after transfer of genetically engineered lymphocytes. Science 314 , 126-129.

Morgan, R.A., Johnson, L.A., Davis, J.L., Zheng, Z., Woolard, K.D., Reap, E.A., Feldman, S.A., Chinnasamy, N., Kuan, C.T., Song, H., et al. (2012). Recognition of glioma stem cells by genetically modified T cells targeting EGFRvIII and development of adoptive cell therapy for glioma. Human gene therapy 23 , 1043-1053.

Morgan, R.A., Yang, J.C., Kitano, M., Dudley, M.E., Laurencot, C.M., and Rosenberg, S.A. (2010). Case report of a serious adverse event following the administration of T cells transduced with a chimeric antigen receptor recognizing ERBB2. Molecular therapy : the journal of the American Society of Gene Therapy 18 , 843-851.

Motz, G.T., Santoro, S.P., Wang, L.P., Garrabrant, T., Lastra, R.R., Hagemann, I.S., Lal, P., Feldman, M.D., Benencia, F., and Coukos, G. (2014). Tumor endothelium FasL establishes a selective immune barrier promoting tolerance in tumors. Nat Med 20 , 607- 615.

Moudgil, K.D., and Choubey, D. (2011). Cytokines in autoimmunity: role in induction, regulation, and treatment. J Interferon Cytokine Res 31 , 695-703.

Mueller, S.N., Gebhardt, T., Carbone, F.R., and Heath, W.R. (2013). Memory T cell subsets, migration patterns, and tissue residence. Annual review of immunology 31 , 137- 161.

Nagata, S., and Golstein, P. (1995). The Fas death factor. Science 267 , 1449-1456.

Naylor, K., Li, G., Vallejo, A.N., Lee, W.W., Koetz, K., Bryl, E., Witkowski, J., Fulbright, J., Weyand, C.M., and Goronzy, J.J. (2005). The influence of age on T cell generation and TCR diversity. J Immunol 174 , 7446-7452.

210

Neely, G.G., Robbins, S.M., Amankwah, E.K., Epelman, S., Wong, H., Spurrell, J.C., Jandu, K.K., Zhu, W., Fogg, D.K., Brown, C.B., and Mody, C.H. (2001). Lipopolysaccharide-stimulated or granulocyte-macrophage colony-stimulating factor- stimulated monocytes rapidly express biologically active IL-15 on their cell surface independent of new protein synthesis. J Immunol 167 , 5011-5017.

Nieminen, A.L., Petrie, T.G., Lemasters, J.J., and Selman, W.R. (1996). Cyclosporin A delays mitochondrial depolarization induced by N-methyl-D-aspartate in cortical neurons: evidence of the mitochondrial permeability transition. Neuroscience 75 , 993- 997.

Nomi, T., Sho, M., Akahori, T., Hamada, K., Kubo, A., Kanehiro, H., Nakamura, S., Enomoto, K., Yagita, H., Azuma, M., and Nakajima, Y. (2007). Clinical significance and therapeutic potential of the programmed death-1 ligand/programmed death-1 pathway in human pancreatic cancer. Clin Cancer Res 13 , 2151-2157.

Novy, P., Huang, X., Leonard, W.J., and Yang, Y. (2011). Intrinsic IL-21 signaling is critical for CD8 T cell survival and memory formation in response to vaccinia viral infection. J Immunol 186 , 2729-2738.

O'Day, S.J., Maio, M., Chiarion-Sileni, V., Gajewski, T.F., Pehamberger, H., Bondarenko, I.N., Queirolo, P., Lundgren, L., Mikhailov, S., Roman, L. , et al. (2010). Efficacy and safety of ipilimumab monotherapy in patients with pretreated advanced melanoma: a multicenter single-arm phase II study. Annals of oncology : official journal of the European Society for Medical Oncology / ESMO 21 , 1712-1717.

Obad, S., dos Santos, C.O., Petri, A., Heidenblad, M., Broom, O., Ruse, C., Fu, C., Lindow, M., Stenvang, J., Straarup, E.M. , et al. (2011). Silencing of microRNA families by seed-targeting tiny LNAs. Nat Genet 43 , 371-378.

Obad S, d.S.C., Petri A, Heidenblad M, Broom O, Ruse C, Fu C, Lindow M, Stenvang J, Straarup EM, Hansen HF, Koch T, Pappin D, Hannon GJ, Kauppinen S. (2011). Silencing of microRNA families by seed-targeting tiny LNAs. Nat Genet 43 , 371-378.

Ohigashi, Y., Sho, M., Yamada, Y., Tsurui, Y., Hamada, K., Ikeda, N., Mizuno, T., Yoriki, R., Kashizuka, H., Yane, K. , et al. (2005). Clinical significance of programmed death-1 ligand-1 and programmed death-1 ligand-2 expression in human esophageal cancer. Clin Cancer Res 11 , 2947-2953.

211

Ohishi, K., Katayama, N., Shiku, H., Varnum-Finney, B., and Bernstein, I.D. (2003). Notch signalling in hematopoiesis. Seminars in cell & developmental biology 14 , 143- 150.

Olson, J.A., McDonald-Hyman, C., Jameson, S.C., and Hamilton, S.E. (2013). Effector- like CD8(+) T cells in the memory population mediate potent protective immunity. Immunity 38 , 1250-1260.

Orom, U.A., Kauppinen, S., and Lund, A.H. (2006). LNA-modified oligonucleotides mediate specific inhibition of microRNA function. Gene 372 , 137-141.

Overwijk, W.W., Theoret, M.R., Finkelstein, S.E., Surman, D.R., de Jong, L.A., Vyth- Dreese, F.A., Dellemijn, T.A., Antony, P.A., Spiess, P.J., Palmer, D.C. , et al. (2003). Tumor regression and autoimmunity after reversal of a functionally tolerant state of self- reactive CD8+ T cells. J Exp Med 198 , 569-580.

Palmer, D.C., Chan, C.C., Gattinoni, L., Wrzesinski, C., Paulos, C.M., Hinrichs, C.S., Powell, D.J., Jr., Klebanoff, C.A., Finkelstein, S.E., Fariss, R.N. , et al. (2008). Effective tumor treatment targeting a melanoma/melanocyte-associated antigen triggers severe ocular autoimmunity. Proc Natl Acad Sci U S A 105 , 8061-8066.

Park, T.S., Rosenberg, S.A., and Morgan, R.A. (2011). Treating cancer with genetically engineered T cells. Trends Biotechnol 29 , 550-557.

Pearce EL, M.A., Martins GA, Krawczyk CM, Hutchins AS, Zediak VP, Banica M, DiCioccio CB, Gross DA, Mao CA, Shen H, Cereb N, Yang SY, Lindsten T, Rossant J, Hunter CA, Reiner SL. (2003). Control of effector CD8+ T cell function by the transcription factor Eomesodermin. Science 302 , 1041-1043.

Pedroza-Pacheco, I., Madrigal, A., and Saudemont, A. (2013). Interaction between natural killer cells and regulatory T cells: perspectives for immunotherapy. Cell Mol Immunol 10 , 222-229.

Peixoto, A., Evaristo, C., Munitic, I., Monteiro, M., Charbit, A., Rocha, B., and Veiga- Fernandes, H. (2007). CD8 single-cell gene coexpression reveals three different effector types present at distinct phases of the immune response. J Exp Med 204 , 1193-1205.

212

Pellegrini, M., Calzascia, T., Elford, A.R., Shahinian, A., Lin, A.E., Dissanayake, D., Dhanji, S., Nguyen, L.T., Gronski, M.A., Morre, M. , et al. (2009). Adjuvant IL-7 antagonizes multiple cellular and molecular inhibitory networks to enhance immunotherapies. Nat Med 15 , 528-536.

Penafuerte, C., and Galipeau, J. (2008). TGF beta secreted by B16 melanoma antagonizes cancer gene immunotherapy bystander effect. Cancer Immunol Immunother 57 , 1197- 1206.

Pender, M.P. (2007). Treating autoimmune demyelination by augmenting lymphocyte apoptosis in the central nervous system. J Neuroimmunol 191 , 26-38.

Perna, S.K., Pagliara, D., Mahendravada, A., Liu, H., Brenner, M.K., Savoldo, B., and Dotti, G. (2014). Interleukin-7 mediates selective expansion of tumor-redirected cytotoxic T lymphocytes (CTLs) without enhancement of regulatory T-cell inhibition. Clin Cancer Res 20 , 131-139.

Petrella, T.M., Tozer, R., Belanger, K., Savage, K.J., Wong, R., Smylie, M., Kamel-Reid, S., Tron, V., Chen, B.E., Hunder, N.N. , et al. (2012). Interleukin-21 has activity in patients with metastatic melanoma: a phase II study. J Clin Oncol 30 , 3396-3401.

Pietenpol, J.A., Stein, R.W., Moran, E., Yaciuk, P., Schlegel, R., Lyons, R.M., Pittelkow, M.R., Munger, K., Howley, P.M., and Moses, H.L. (1990). TGF-beta 1 inhibition of c- myc transcription and growth in keratinocytes is abrogated by viral transforming proteins with pRB binding domains. Cell 61 , 777-785.

Pilon-Thomas S, K.L., Ellwanger S, Janssen W, Royster E, Marzban S, Kudchadkar R, Zager J, Gibney G, Sondak VK, Weber J, Mulé JJ, Sarnaik AA (2012). Efficacy of adoptive cell transfer of tumor-infiltrating lymphocytes after lymphopenia induction for metastatic melanoma. J Immunother 35 , 615-620.

Polak ME, B.N., Gabriel FG, Johnson P, Higgins B, Hurren J, McCormick D, Jager MJ, Cree IA (2007). Mechanisms of local immunosuppression in cutaneous melanoma. Br J Cancer 96 , 1879-1887.

Powell, D.J., Jr., de Vries, C.R., Allen, T., Ahmadzadeh, M., and Rosenberg, S.A. (2007). Inability to mediate prolonged reduction of regulatory T Cells after transfer of autologous

213

CD25-depleted PBMC and interleukin-2 after lymphodepleting chemotherapy. J Immunother 30 , 438-447.

Prado-Garcia, H., Aguilar-Cazares, D., Flores-Vergara, H., Mandoki, J.J., and Lopez- Gonzalez, J.S. (2005). Effector, memory and naive CD8+ T cells in peripheral blood and pleural effusion from lung adenocarcinoma patients. Lung Cancer 47 , 361-371.

Prado-Garcia, H., Romero-Garcia, S., Aguilar-Cazares, D., Meneses-Flores, M., and Lopez-Gonzalez, J.S. (2012). Tumor-induced CD8+ T-cell dysfunction in lung cancer patients. Clin Dev Immunol 2012 , Article ID 741741.

Quan, L.T., Tewari, M., O'Rourke, K., Dixit, V., Snipas, S.J., Poirier, G.G., Ray, C., Pickup, D.J., and Salvesen, G.S. (1996). Proteolytic activation of the cell death protease Yama/CPP32 by granzyme B. Proc Natl Acad Sci U S A 93 , 1972-1976.

Quezada, S.A., Peggs, K.S., Simpson, T.R., and Allison, J.P. (2011). Shifting the equilibrium in cancer immunoediting: from tumor tolerance to eradication. Immunological reviews 241 , 104-118.

Qureshi, O.S., Zheng, Y., Nakamura, K., Attridge, K., Manzotti, C., Schmidt, E.M., Baker, J., Jeffery, L.E., Kaur, S., Briggs, Z. , et al. (2011). Trans-endocytosis of CD80 and CD86: a molecular basis for the cell-extrinsic function of CTLA-4. Science 332 , 600- 603.

Radvanyi, L.G., Bernatchez, C., Zhang, M., Fox, P.S., Miller, P., Chacon, J., Wu, R., Lizee, G., Mahoney, S., Alvarado, G. , et al. (2012). Specific lymphocyte subsets predict response to adoptive cell therapy using expanded autologous tumor-infiltrating lymphocytes in metastatic melanoma patients. Clin Cancer Res 18 , 6758-6770.

Rao, R.R., Li, Q., Odunsi, K., and Shrikant, P.A. (2010). The mTOR kinase determines effector versus memory CD8+ T cell fate by regulating the expression of transcription factors T-bet and Eomesodermin. Immunity 32 , 67-78.

Remitz, A., Lauerma, A.I., Erkko, P., and Reitamo, S. (1996). Delayed-type hypersensitivity in palmoplantar pustulosis: effect of cyclosporin A treatment on skin testing with recall antigens. Acta dermato-venereologica 76 , 310-313.

214

Ribas, A., Kefford, R., Marshall, M.A., Punt, C.J., Haanen, J.B., Marmol, M., Garbe, C., Gogas, H., Schachter, J., Linette, G. , et al. (2013). Phase III randomized clinical trial comparing tremelimumab with standard-of-care chemotherapy in patients with advanced melanoma. J Clin Oncol 31 , 616-622.

Riley, J.L., Mao, M., Kobayashi, S., Biery, M., Burchard, J., Cavet, G., Gregson, B.P., June, C.H., and Linsley, P.S. (2002). Modulation of TCR-induced transcriptional profiles by ligation of CD28, ICOS, and CTLA-4 receptors. Proc Natl Acad Sci U S A 99 , 11790- 11795.

Robbins PF, M.R., Feldman SA, Yang JC, Sherry RM, Dudley ME, Wunderlich JR, Nahvi AV, Helman LJ, Mackall CL, Kammula US, Hughes MS, Restifo NP, Raffeld M, Lee CC, Levy CL, Li YF, El-Gamil M, Schwarz SL, Laurencot C, Rosenberg SA (2011). Tumor regression in patients with metastatic synovial cell sarcoma and melanoma using genetically engineered lymphocytes reactive with NY-ESO-1. J Clin Oncol 29 , 917-924.

Robert, C., Long, G.V., Brady, B., Dutriaux, C., Maio, M., Mortier, L., Hassel, J.C., Rutkowski, P., McNeil, C., Kalinka-Warzocha, E. , et al. (2015). Nivolumab in previously untreated melanoma without BRAF mutation. N Engl J Med 372 , 320-330.

Rodriguez A, V.E., Clare S, Warren MV, Couttet P, Soond DR, van Dongen S, Grocock RJ, Das PP, Miska EA, Vetrie D, Okkenhaug K, Enright AJ, Dougan G, Turner M, Bradley A. (2007). Requirement of bic/microRNA-155 for normal immune function. Science 316 , 608-611.

Rosenberg, S.A. (2001). Progress in human tumour immunology and immunotherapy. Nature 411 , 380-384.

Rosenberg, S.A., and Dudley, M.E. (2004). Cancer regression in patients with metastatic melanoma after the transfer of autologous antitumor lymphocytes. Proc Natl Acad Sci U S A 101 Suppl 2 , 14639-14645.

Rosenberg, S.A., Restifo, N.P., Yang, J.C., Morgan, R.A., and Dudley, M.E. (2008). Adoptive cell transfer: a clinical path to effective cancer immunotherapy. Nat Rev Cancer 8, 299-308.

Rosenberg SA, R.N., Yang JC, Morgan RA, Dudley ME. (2008). Adoptive cell transfer: a clinical path to effective cancer immunotherapy. Nat Rev Cancer. 8, 299-308. 215

Rosenberg, S.A., Yang, J.C., Sherry, R.M., Kammula, U.S., Hughes, M.S., Phan, G.Q., Citrin, D.E., Restifo, N.P., Robbins, P.F., Wunderlich, J.R. , et al. (2011). Durable complete responses in heavily pretreated patients with metastatic melanoma using T-cell transfer immunotherapy. Clin Cancer Res 17 , 4550-4557.

Rosenberg SA, Y.J., Sherry RM, Kammula US, Hughes MS, Phan GQ, Citrin DE, Restifo NP, Robbins PF, Wunderlich JR, Morton KE, Laurencot CM, Steinberg SM, White DE, Dudley ME (2011). Durable complete responses in heavily pretreated patients with metastatic melanoma using T-cell transfer immunotherapy. Clin Cancer Res 17 , 4550-4557.

Rudd, B.D., Venturi, V., Li, G., Samadder, P., Ertelt, J.M., Way, S.S., Davenport, M.P., and Nikolich-Zugich, J. (2011). Nonrandom attrition of the naive CD8+ T-cell pool with aging governed by T-cell receptor:pMHC interactions. Proc Natl Acad Sci U S A 108 , 13694-13699.

Ruegemer, J.J., Ho, S.N., Augustine, J.A., Schlager, J.W., Bell, M.P., McKean, D.J., and Abraham, R.T. (1990). Regulatory effects of transforming growth factor-beta on IL-2- and IL-4-dependent T cell-cycle progression. J Immunol 144 , 1767-1776.

Rutishauser, R.L., Martins, G.A., Kalachikov, S., Chandele, A., Parish, I.A., Meffre, E., Jacob, J., Calame, K., and Kaech, S.M. (2009). Transcriptional repressor Blimp-1 promotes CD8(+) T cell terminal differentiation and represses the acquisition of central memory T cell properties. Immunity 31 , 296-308.

Rutishauser RL, M.G., Kalachikov S, Chandele A, Parish IA, Meffre E, Jacob J, Calame K, Kaech SM. (2009). Transcriptional repressor Blimp-1 promotes CD8(+) T cell terminal differentiation and represses the acquisition of central memory T cell properties. Immunity. 31 , 296-308.

Saint-Mezard, P., Berard, F., Dubois, B., Kaiserlian, D., and Nicolas, J.F. (2004). The role of CD4+ and CD8+ T cells in contact hypersensitivity and allergic contact dermatitis. Eur J Dermatol 14 , 131-138.

Saligrama, P.T., Fortner, K.A., Secinaro, M.A., Collins, C.C., Russell, J.Q., and Budd, R.C. (2014). IL-15 maintains T-cell survival via S-nitrosylation-mediated inhibition of caspase-3. Cell Death Differ 21 , 904-914.

216

Sallusto, F., Lenig, D., Forster, R., Lipp, M., and Lanzavecchia, A. (1999). Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401 , 708-712.

Sampson, J.H., Choi, B.D., Sanchez-Perez, L., Suryadevara, C.M., Snyder, D.J., Flores, C.T., Schmittling, R.J., Nair, S.K., Reap, E.A., Norberg, P.K. , et al. (2014). EGFRvIII mCAR-modified T-cell therapy cures mice with established intracerebral glioma and generates host immunity against tumor-antigen loss. Clin Cancer Res 20 , 972-984.

Sathish, J.G., Sethu, S., Bielsky, M.C., de Haan, L., French, N.S., Govindappa, K., Green, J., Griffiths, C.E., Holgate, S., Jones, D. , et al. (2013). Challenges and approaches for the development of safer immunomodulatory biologics. Nat Rev Drug Discov 12 , 306-324.

Sauter, B., Albert, M.L., Francisco, L., Larsson, M., Somersan, S., and Bhardwaj, N. (2000). Consequences of cell death: exposure to necrotic tumor cells, but not primary tissue cells or apoptotic cells, induces the maturation of immunostimulatory dendritic cells. J Exp Med 191 , 423-434.

Schenkel, J.M., Fraser, K.A., Beura, L.K., Pauken, K.E., Vezys, V., and Masopust, D. (2014). T cell memory. Resident memory CD8 T cells trigger protective innate and adaptive immune responses. Science 346 , 98-101.

Schenkel, J.M., Fraser, K.A., Vezys, V., and Masopust, D. (2013). Sensing and alarm function of resident memory CD8(+) T cells. Nat Immunol 14 , 509-513.

Schneider, H., Mandelbrot, D.A., Greenwald, R.J., Ng, F., Lechler, R., Sharpe, A.H., and Rudd, C.E. (2002). Cutting edge: CTLA-4 (CD152) differentially regulates mitogen- activated protein kinases (extracellular signal-regulated kinase and c-Jun N-terminal kinase) in CD4+ T cells from receptor/ligand-deficient mice. J Immunol 169 , 3475-3479.

Scholler, J., Brady, T.L., Binder-Scholl, G., Hwang, W.T., Plesa, G., Hege, K.M., Vogel, A.N., Kalos, M., Riley, J.L., Deeks, S.G. , et al. (2012). Decade-long safety and function of retroviral-modified chimeric antigen receptor T cells. Sci Transl Med 4, 132ra153.

Schwarz, K., van Den Broek, M., Kostka, S., Kraft, R., Soza, A., Schmidtke, G., Kloetzel, P.M., and Groettrup, M. (2000). Overexpression of the proteasome subunits LMP2, LMP7, and MECL-1, but not PA28 alpha/beta, enhances the presentation of an 217

immunodominant lymphocytic choriomeningitis virus T cell epitope. J Immunol 165 , 768-778.

Seddiki, N., Santner-Nanan, B., Martinson, J., Zaunders, J., Sasson, S., Landay, A., Solomon, M., Selby, W., Alexander, S.I., Nanan, R. , et al. (2006). Expression of interleukin (IL)-2 and IL-7 receptors discriminates between human regulatory and activated T cells. J Exp Med 203 , 1693-1700.

Seliger, B., Hammers, S., Hohne, A., Zeidler, R., Knuth, A., Gerharz, C.D., and Huber, C. (1997). IFN-gamma-mediated coordinated transcriptional regulation of the human TAP-1 and LMP-2 genes in human renal cell carcinoma. Clin Cancer Res 3, 573-578.

Sena, L.A., Li, S., Jairaman, A., Prakriya, M., Ezponda, T., Hildeman, D.A., Wang, C.R., Schumacker, P.T., Licht, J.D., Perlman, H. , et al. (2013). Mitochondria are required for antigen-specific T cell activation through reactive oxygen species signaling. Immunity 38 , 225-236.

Sheppard, H.M., Verdon, D., Brooks, A., Feisst, V., Ho, Y.Y., Lorenz, N., Fan, V., Birch, N.P., Didsbury, A., and Dunbar, P. (2014). MicroRNA regulation in human CD8+ T cell subsets inverted question mark cytokine exposure alone drives miR-146a expression. J Transl Med 12 , 292.

Shin, H., Blackburn, S.D., Intlekofer, A.M., Kao, C., Angelosanto, J.M., Reiner, S.L., and Wherry, E.J. (2009). A role for the transcriptional repressor Blimp-1 in CD8(+) T cell exhaustion during chronic viral infection. Immunity 31 , 309-320.

Shin H, B.S., Intlekofer AM, Kao C, Angelosanto JM, Reiner SL, Wherry EJ. (2009). A role for the transcriptional repressor Blimp-1 in CD8(+) T cell exhaustion during chronic viral infection. Immunity. 31 , 309-320.

Shin, H.M., Kapoor, V.N., Guan, T., Kaech, S.M., Welsh, R.M., and Berg, L.J. (2013). Epigenetic modifications induced by Blimp-1 Regulate CD8(+) T cell memory progression during acute virus infection. Immunity 39 , 661-675.

Shin, M.S., Lee, N., and Kang, I. (2011). Effector T-cell subsets in systemic lupus erythematosus: update focusing on Th17 cells. Curr Opin Rheumatol 23 , 444-448.

218

Sikora, J.J., Dworacki, G.T., Kaczmarek, M.T., Jenek, R.E., and Zeromski, J.O. (2004). Immunosuppressive mechanisms in the microenvironment of malignant pleural effusions. Cancer Detect Prev 28 , 325-330.

Sim, G.C., Martin-Orozco, N., Jin, L., Yang, Y., Wu, S., Washington, E., Sanders, D., Lacey, C., Wang, Y., Vence, L. , et al. (2014). IL-2 therapy promotes suppressive ICOS+ Treg expansion in melanoma patients. J Clin Invest 124 , 99-110.

Singh, R.P., Hasan, S., Sharma, S., Nagra, S., Yamaguchi, D.T., Wong, D.T., Hahn, B.H., and Hossain, A. (2014). Th17 cells in inflammation and autoimmunity. Autoimmun Rev 13 , 1174-1181.

Smyth, M.J., Cretney, E., Kelly, J.M., Westwood, J.A., Street, S.E., Yagita, H., Takeda, K., van Dommelen, S.L., Degli-Esposti, M.A., and Hayakawa, Y. (2005). Activation of NK cell cytotoxicity. Mol Immunol 42 , 501-510.

Smyth, M.J., Thia, K.Y., Street, S.E., MacGregor, D., Godfrey, D.I., and Trapani, J.A. (2000). Perforin-mediated cytotoxicity is critical for surveillance of spontaneous lymphoma. J Exp Med 192 , 755-760.

Snyder, A., Makarov, V., Merghoub, T., Yuan, J., Zaretsky, J.M., Desrichard, A., Walsh, L.A., Postow, M.A., Wong, P., Ho, T.S. , et al. (2014). Genetic basis for clinical response to CTLA-4 blockade in melanoma. N Engl J Med 371 , 2189-2199.

Sorensen, P.S., Bertolotto, A., Edan, G., Giovannoni, G., Gold, R., Havrdova, E., Kappos, L., Kieseier, B.C., Montalban, X., and Olsson, T. (2012). Risk stratification for progressive multifocal leukoencephalopathy in patients treated with natalizumab. Multiple sclerosis 18 , 143-152.

Spear, P., Barber, A., Rynda-Apple, A., and Sentman, C.L. (2012). Chimeric antigen receptor T cells shape myeloid cell function within the tumor microenvironment through IFN-gamma and GM-CSF. J Immunol 188 , 6389-6398.

Spear, P., Barber, A., Rynda-Apple, A., and Sentman, C.L. (2013a). NKG2D CAR T-cell therapy inhibits the growth of NKG2D ligand heterogeneous tumors. Immunology and cell biology 91 , 435-440.

219

Spear, P., Barber, A., and Sentman, C.L. (2013b). Collaboration of chimeric antigen receptor (CAR)-expressing T cells and host T cells for optimal elimination of established ovarian tumors. Oncoimmunology 2, e23564.

Sportes, C., Babb, R.R., Krumlauf, M.C., Hakim, F.T., Steinberg, S.M., Chow, C.K., Brown, M.R., Fleisher, T.A., Noel, P., Maric, I. , et al. (2010). Phase I study of recombinant human interleukin-7 administration in subjects with refractory malignancy. Clin Cancer Res 16 , 727-735.

Sportes, C., Hakim, F.T., Memon, S.A., Zhang, H., Chua, K.S., Brown, M.R., Fleisher, T.A., Krumlauf, M.C., Babb, R.R., Chow, C.K. , et al. (2008). Administration of rhIL-7 in humans increases in vivo TCR repertoire diversity by preferential expansion of naive T cell subsets. J Exp Med 205 , 1701-1714.

Steele, N., Anthony, A., Saunders, M., Esmarck, B., Ehrnrooth, E., Kristjansen, P.E., Nihlen, A., Hansen, L.T., and Cassidy, J. (2012). A phase 1 trial of recombinant human IL-21 in combination with cetuximab in patients with metastatic colorectal cancer. Br J Cancer 106 , 793-798.

Stephen, T.L., Rutkowski, M.R., Allegrezza, M.J., Perales-Puchalt, A., Tesone, A.J., Svoronos, N., Nguyen, J.M., Sarmin, F., Borowsky, M.E., Tchou, J., and Conejo-Garcia, J.R. (2014). Transforming growth factor beta-mediated suppression of antitumor T cells requires FoxP1 transcription factor expression. Immunity 41 , 427-439.

Stepkowski, S.M. (2000). Molecular targets for existing and novel immunosuppressive drugs. Expert Rev Mol Med 2, 1-23.

Stoklasek, T.A., Schluns, K.S., and Lefrancois, L. (2006). Combined IL-15/IL-15Ralpha immunotherapy maximizes IL-15 activity in vivo. J Immunol 177 , 6072-6080.

Street, S.E., Cretney, E., and Smyth, M.J. (2001). Perforin and interferon-gamma activities independently control tumor initiation, growth, and metastasis. Blood 97 , 192- 197.

Strobel, G.A., and Pliam, N.B. (1996). Immunosuppressant secondary metabolite diterpenes and method of production using a fungal organism. (Application: WO1996022382 A1, Pharmagenesis, Inc., USA.), p. 45 pp.

220

Stupp, R., Mason, W.P., van den Bent, M.J., Weller, M., Fisher, B., Taphoorn, M.J., Belanger, K., Brandes, A.A., Marosi, C., Bogdahn, U. , et al. (2005). Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 352 , 987-996.

Sugawa, N., Ekstrand, A.J., James, C.D., and Collins, V.P. (1990). Identical splicing of aberrant epidermal growth factor receptor transcripts from amplified rearranged genes in human glioblastomas. Proc Natl Acad Sci U S A 87 , 8602-8606.

Sugimoto, K., Maekawa, Y., Kitamura, A., Nishida, J., Koyanagi, A., Yagita, H., Kojima, H., Chiba, S., Shimada, M., and Yasutomo, K. (2010). Notch2 signaling is required for potent antitumor immunity in vivo. J Immunol 184 , 4673-4678.

Sullivan BM, J.A., Szabo SJ, von Herrath M, Glimcher LH. (2003). Antigen-driven effector CD8 T cell function regulated by T-bet. Proc Natl Acad Sci U S A. 100 , 15818- 15823.

Szabo, S.J., Kim, S.T., Costa, G.L., Zhang, X., Fathman, C.G., and Glimcher, L.H. (2000). A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 100 , 655-669.

Tajima, K., Ito, Y., Demachi, A., Nishida, K., Akatsuka, Y., Tsujimura, K., Hida, T., Morishima, Y., Kuwano, H., Mitsudomi, T. , et al. (2004). Interferon-gamma differentially regulates susceptibility of lung cancer cells to telomerase-specific cytotoxic T lymphocytes. Int J Cancer 110 , 403-412.

Tao, X., and Lipsky, P.E. (2000). The Chinese anti-inflammatory and immunosuppressive herbal remedy Tripterygium wilfordii Hook F. Rheum Dis Clin North Am 26 , 29-50, viii.

Tchervenkov, J.I., Diano, E., and Christou, N.V. (1985). Effect of cyclosporine on host defence. Canadian journal of surgery. Journal canadien de chirurgie 28 , 525-527.

Terabe, M., Ambrosino, E., Takaku, S., O'Konek, J.J., Venzon, D., Lonning, S., McPherson, J.M., and Berzofsky, J.A. (2009). Synergistic enhancement of CD8+ T cell- mediated tumor vaccine efficacy by an anti-transforming growth factor-beta monoclonal antibody. Clin Cancer Res 15 , 6560-6569.

221

Terme, M., Ullrich, E., Aymeric, L., Meinhardt, K., Desbois, M., Delahaye, N., Viaud, S., Ryffel, B., Yagita, H., Kaplanski, G. , et al. (2011). IL-18 induces PD-1-dependent immunosuppression in cancer. Cancer Res 71 , 5393-5399.

Thai TH, C.D., Casola S, Ansel KM, Xiao C, Xue Y, Murphy A, Frendewey D, Valenzuela D, Kutok JL, Schmidt-Supprian M, Rajewsky N, Yancopoulos G, Rao A, Rajewsky K. (2007 ). Regulation of the germinal center response by microRNA-155. Science 316 , 604-608.

The Autoimmune Diseases Coordinating Committee, N. (2005). Progress in autoimmune disease research, Report to Congress. (National Institute of Health).

Thomas, D.A., and Massague, J. (2005). TGF-beta directly targets cytotoxic T cell functions during tumor evasion of immune surveillance. Cancer Cell 8, 369-380.

Thomas DA, M.J. (2005). TGF-beta directly targets cytotoxic T cell functions during tumor evasion of immune surveillance. Cancer Cell 8, 369-380.

Thompson, J.A., Curti, B.D., Redman, B.G., Bhatia, S., Weber, J.S., Agarwala, S.S., Sievers, E.L., Hughes, S.D., DeVries, T.A., and Hausman, D.F. (2008). Phase I study of recombinant interleukin-21 in patients with metastatic melanoma and renal cell carcinoma. J Clin Oncol 26 , 2034-2039.

Thum T, G.C., Fiedler J, Fischer T, Kissler S, Bussen M, Galuppo P, Just S, Rottbauer W, Frantz S, Castoldi M, Soutschek J, Koteliansky V, Rosenwald A, Basson MA, Licht JD, Pena JT, Rouhanifard SH, Muckenthaler MU, Tuschl T, Martin GR, Bauersachs J, Engelhardt S. (2008). MicroRNA-21 contributes to myocardial disease by stimulating MAP kinase signalling in fibroblasts. Nature 456 , 980-984.

Till, B.G., Jensen, M.C., Wang, J., Chen, E.Y., Wood, B.L., Greisman, H.A., Qian, X., James, S.E., Raubitschek, A., Forman, S.J. , et al. (2008). Adoptive immunotherapy for indolent non-Hodgkin lymphoma and mantle cell lymphoma using genetically modified autologous CD20-specific T cells. Blood 112 , 2261-2271.

Toes, R.E., Nussbaum, A.K., Degermann, S., Schirle, M., Emmerich, N.P., Kraft, M., Laplace, C., Zwinderman, A., Dick, T.P., Muller, J., et al. (2001). Discrete cleavage motifs of constitutive and immunoproteasomes revealed by quantitative analysis of cleavage products. J Exp Med 194 , 1-12. 222

Tone, Y., Furuuchi, K., Kojima, Y., Tykocinski, M.L., Greene, M.I., and Tone, M. (2008). Smad3 and NFAT cooperate to induce Foxp3 expression through its enhancer. Nat Immunol 9, 194-202.

Topalian, S.L., Hodi, F.S., Brahmer, J.R., Gettinger, S.N., Smith, D.C., McDermott, D.F., Powderly, J.D., Carvajal, R.D., Sosman, J.A., Atkins, M.B. , et al. (2012). Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. N Engl J Med 366 , 2443-2454.

Tran, E., Turcotte, S., Gros, A., Robbins, P.F., Lu, Y.C., Dudley, M.E., Wunderlich, J.R., Somerville, R.P., Hogan, K., Hinrichs, C.S. , et al. (2014). Cancer immunotherapy based on mutation-specific CD4+ T cells in a patient with epithelial cancer. Science 344 , 641- 645.

Triebel, F., Jitsukawa, S., Baixeras, E., Roman-Roman, S., Genevee, C., Viegas- Pequignot, E., and Hercend, T. (1990). LAG-3, a novel lymphocyte activation gene closely related to CD4. J Exp Med 171 , 1393-1405.

Trifari, S., Pipkin, M.E., Bandukwala, H.S., Aijo, T., Bassein, J., Chen, R., Martinez, G.J., and Rao, A. (2013). MicroRNA-directed program of cytotoxic CD8+ T-cell differentiation. Proc Natl Acad Sci U S A 110 , 18608-18613.

Tumeh, P.C., Harview, C.L., Yearley, J.H., Shintaku, I.P., Taylor, E.J., Robert, L., Chmielowski, B., Spasic, M., Henry, G., Ciobanu, V., et al. (2014). PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 515 , 568-571.

Turner, C.A., Jr., Mack, D.H., and Davis, M.M. (1994). Blimp-1, a novel zinc finger- containing protein that can drive the maturation of B lymphocytes into immunoglobulin- secreting cells. Cell 77 , 297-306.

Ueda, R., Fujita, M., Zhu, X., Sasaki, K., Kastenhuber, E.R., Kohanbash, G., McDonald, H.A., Harper, J., Lonning, S., and Okada, H. (2009). Systemic inhibition of transforming growth factor-beta in glioma-bearing mice improves the therapeutic efficacy of glioma- associated antigen peptide vaccines. Clin Cancer Res 15 , 6551-6559.

Van Belle P, R.U., Nuamah I, Halpern AC, Elder DE (1996). Melanoma-associated expression of transforming growth factor-beta isoforms. Am J Pathol 148 , 1887-1894.

223

van der Bruggen, P., Traversari, C., Chomez, P., Lurquin, C., De Plaen, E., Van den Eynde, B., Knuth, A., and Boon, T. (1991). A gene encoding an antigen recognized by cytolytic T lymphocytes on a human melanoma. Science 254 , 1643-1647. van der Windt, G.J., Everts, B., Chang, C.H., Curtis, J.D., Freitas, T.C., Amiel, E., Pearce, E.J., and Pearce, E.L. (2012). Mitochondrial respiratory capacity is a critical regulator of CD8+ T cell memory development. Immunity 36 , 68-78. van Elsas, A., Hurwitz, A.A., and Allison, J.P. (1999). Combination immunotherapy of B16 melanoma using anti-cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) and granulocyte/macrophage colony-stimulating factor (GM-CSF)-producing vaccines induces rejection of subcutaneous and metastatic tumors accompanied by autoimmune depigmentation. J Exp Med 190 , 355-366.

Venturi, V., Price, D.A., Douek, D.C., and Davenport, M.P. (2008). The molecular basis for public T-cell responses? Nat Rev Immunol 8, 231-238.

Viard-Leveugle, I., Veyrenc, S., French, L.E., Brambilla, C., and Brambilla, E. (2003). Frequent loss of Fas expression and function in human lung tumours with overexpression of FasL in small cell lung carcinoma. The Journal of pathology 201 , 268-277.

Vigano, S., Perreau, M., Pantaleo, G., and Harari, A. (2012). Positive and negative regulation of cellular immune responses in physiologic conditions and diseases. Clin Dev Immunol 2012 , 485781.

Vilches, C., and Parham, P. (2002). KIR: diverse, rapidly evolving receptors of innate and adaptive immunity. Annual review of immunology 20 , 217-251.

Vitale, M., Cantoni, C., Pietra, G., Mingari, M.C., and Moretta, L. (2014). Effect of tumor cells and tumor microenvironment on NK-cell function. Eur J Immunol 44 , 1582- 1592.

Vocanson, M., Hennino, A., Rozieres, A., Poyet, G., and Nicolas, J.F. (2009). Effector and regulatory mechanisms in allergic contact dermatitis. Allergy 64 , 1699-1714.

Vogelstein, B., Papadopoulos, N., Velculescu, V.E., Zhou, S., Diaz, L.A., Jr., and Kinzler, K.W. (2013). Cancer genome landscapes. Science 339 , 1546-1558. 224

Wan, C.K., Oh, J., Li, P., West, E.E., Wong, E.A., Andraski, A.B., Spolski, R., Yu, Z.X., He, J., Kelsall, B.L., and Leonard, W.J. (2013). The cytokines IL-21 and GM-CSF have opposing regulatory roles in the apoptosis of conventional dendritic cells. Immunity 38 , 514-527.

Wang, R., Dillon, C.P., Shi, L.Z., Milasta, S., Carter, R., Finkelstein, D., McCormick, L.L., Fitzgerald, P., Chi, H., Munger, J., and Green, D.R. (2011). The transcription factor Myc controls metabolic reprogramming upon T lymphocyte activation. Immunity 35 , 871-882.

Waterhouse, N.J., Sutton, V.R., Sedelies, K.A., Ciccone, A., Jenkins, M., Turner, S.J., Bird, P.I., and Trapani, J.A. (2006). Cytotoxic T lymphocyte-induced killing in the absence of granzymes A and B is unique and distinct from both apoptosis and perforin- dependent lysis. The Journal of cell biology 173 , 133-144.

Waterhouse, P., Penninger, J.M., Timms, E., Wakeham, A., Shahinian, A., Lee, K.P., Thompson, C.B., Griesser, H., and Mak, T.W. (1995). Lymphoproliferative disorders with early lethality in mice deficient in Ctla-4. Science 270 , 985-988.

Webb, J.R., Milne, K., Watson, P., Deleeuw, R.J., and Nelson, B.H. (2014). Tumor- infiltrating lymphocytes expressing the tissue resident memory marker CD103 are associated with increased survival in high-grade serous ovarian cancer. Clin Cancer Res 20 , 434-444.

Weber, F., Byrne, S.N., Le, S., Brown, D.A., Breit, S.N., Scolyer, R.A., and Halliday, G.M. (2005). Transforming growth factor-beta1 immobilises dendritic cells within skin tumours and facilitates tumour escape from the immune system. Cancer Immunol Immunother 54 , 898-906.

Weber, J., Thompson, J.A., Hamid, O., Minor, D., Amin, A., Ron, I., Ridolfi, R., Assi, H., Maraveyas, A., Berman, D. , et al. (2009). A randomized, double-blind, placebo- controlled, phase II study comparing the tolerability and efficacy of ipilimumab administered with or without prophylactic budesonide in patients with unresectable stage III or IV melanoma. Clin Cancer Res 15 , 5591-5598.

Weber, K.S., Li, Q.J., Persaud, S.P., Campbell, J.D., Davis, M.M., and Allen, P.M. (2012). Distinct CD4+ helper T cells involved in primary and secondary responses to infection. Proc Natl Acad Sci U S A 109 , 9511-9516.

225

Wherry, E.J., Teichgraber, V., Becker, T.C., Masopust, D., Kaech, S.M., Antia, R., von Andrian, U.H., and Ahmed, R. (2003). Lineage relationship and protective immunity of memory CD8 T cell subsets. Nat Immunol 4, 225-234.

Wikstrand, C.J., Hale, L.P., Batra, S.K., Hill, M.L., Humphrey, P.A., Kurpad, S.N., McLendon, R.E., Moscatello, D., Pegram, C.N., and Reist, C.J. (1995). Monoclonal antibodies against EGFRvIII are tumor specific and react with breast and lung carcinomas and malignant gliomas. Cancer research 55 , 3140-3148.

Wilkie, S., Picco, G., Foster, J., Davies, D.M., Julien, S., Cooper, L., Arif, S., Mather, S.J., Taylor-Papadimitriou, J., Burchell, J.M., and Maher, J. (2008). Retargeting of human T cells to tumor-associated MUC1: the evolution of a chimeric antigen receptor. J Immunol 180 , 4901-4909.

Wille-Reece, U., Flynn, B.J., Lore, K., Koup, R.A., Miles, A.P., Saul, A., Kedl, R.M., Mattapallil, J.J., Weiss, W.R., Roederer, M., and Seder, R.A. (2006). Toll-like receptor agonists influence the magnitude and quality of memory T cell responses after prime- boost immunization in nonhuman primates. J Exp Med 203 , 1249-1258.

Winter, J., Jung, S., Keller, S., Gregory, R.I., and Diederichs, S. (2009). Many roads to maturity: microRNA biogenesis pathways and their regulation. Nature cell biology 11 , 228-234.

Wolfraim, L.A., Walz, T.M., James, Z., Fernandez, T., and Letterio, J.J. (2004). p21Cip1 and p27Kip1 act in synergy to alter the sensitivity of naive T cells to TGF-beta-mediated G1 arrest through modulation of IL-2 responsiveness. J Immunol 173 , 3093-3102.

Wong, A.J., Ruppert, J.M., Bigner, S.H., Grzeschik, C.H., Humphrey, P.A., Bigner, D.S., and Vogelstein, B. (1992). Structural alterations of the epidermal growth factor receptor gene in human gliomas. Proc Natl Acad Sci U S A 89 , 2965-2969.

Woo, S.R., Fuertes, M.B., Corrales, L., Spranger, S., Furdyna, M.J., Leung, M.Y., Duggan, R., Wang, Y., Barber, G.N., Fitzgerald, K.A. , et al. (2014). STING-Dependent Cytosolic DNA Sensing Mediates Innate Immune Recognition of Immunogenic Tumors. Immunity 41 , 830-842.

Woo SR, T.M., Goldberg MV, Bankoti J, Selby M, Nirschl CJ, Bettini ML, Gravano DM, Vogel P, Liu CL, Tangsombatvisit S, Grosso JF, Netto G, Smeltzer MP, Chaux A, 226

Utz PJ, Workman CJ, Pardoll DM, Korman AJ, Drake CG, Vignali DA (2012). Immune inhibitory molecules LAG-3 and PD-1 synergistically regulate T-cell function to promote tumoral immune escape. Cancer Res 72 , 917-927.

Woo, S.R., Turnis, M.E., Goldberg, M.V., Bankoti, J., Selby, M., Nirschl, C.J., Bettini, M.L., Gravano, D.M., Vogel, P., Liu, C.L. , et al. (2012). Immune inhibitory molecules LAG-3 and PD-1 synergistically regulate T-cell function to promote tumoral immune escape. Cancer Res 72 , 917-927.

Wu, H., Neilson, J.R., Kumar, P., Manocha, M., Shankar, P., Sharp, P.A., and Manjunath, N. (2007). miRNA profiling of naive, effector and memory CD8 T cells. PLoS One 2, e1020.

Wu, J.D., Higgins, L.M., Steinle, A., Cosman, D., Haugk, K., and Plymate, S.R. (2004). Prevalent expression of the immunostimulatory MHC class I chain-related molecule is counteracted by shedding in prostate cancer. J Clin Invest 114 , 560-568.

Xiao C, S.L., Calado DP, Patterson HC, Zhang B, Wang J, Henderson JM, Kutok JL, Rajewsky K. (2008). Lymphoproliferative disease and autoimmunity in mice with increased miR-17-92 expression in lymphocytes. Nat Immunol 9, 405-414.

Xiao, F., Zuo, Z., Cai, G., Kang, S., Gao, X., and Li, T. (2009). miRecords: an integrated resource for microRNA-target interactions. Nucleic Acids Res 37 , D105-110.

Xu, X., Fu, X.Y., Plate, J., and Chong, A.S. (1998). IFN-gamma induces cell growth inhibition by Fas-mediated apoptosis: requirement of STAT1 protein for up-regulation of Fas and FasL expression. Cancer Res 58 , 2832-2837.

Yadav, M., Jhunjhunwala, S., Phung, Q.T., Lupardus, P., Tanguay, J., Bumbaca, S., Franci, C., Cheung, T.K., Fritsche, J., Weinschenk, T. , et al. (2014). Predicting immunogenic tumour mutations by combining mass spectrometry and exome sequencing. Nature 515 , 572-576.

Yager, E.J., Ahmed, M., Lanzer, K., Randall, T.D., Woodland, D.L., and Blackman, M.A. (2008). Age-associated decline in T cell repertoire diversity leads to holes in the repertoire and impaired immunity to influenza virus. J Exp Med 205 , 711-723.

227

Yamazaki, H., Ohba, Y., Tamaoki, N., and Shibuya, M. (1990). A deletion mutation within the ligand binding domain is responsible for activation of epidermal growth factor receptor gene in human brain tumors. Japanese journal of cancer research : Gann 81 , 773- 779.

Yamazaki, T., Akiba, H., Iwai, H., Matsuda, H., Aoki, M., Tanno, Y., Shin, T., Tsuchiya, H., Pardoll, D.M., Okumura, K. , et al. (2002). Expression of programmed death 1 ligands by murine T cells and APC. J Immunol 169 , 5538-5545.

Yang, L., and Baltimore, D. (2005). Long-term in vivo provision of antigen-specific T cell immunity by programming hematopoietic stem cells. Proc Natl Acad Sci U S A 102 , 4518-4523.

Yin, X., Giap, C., Lazo, J.S., and Prochownik, E.V. (2003). Low molecular weight inhibitors of Myc-Max interaction and function. Oncogene 22 , 6151-6159.

Yokosuka, T., Takamatsu, M., Kobayashi-Imanishi, W., Hashimoto-Tane, A., Azuma, M., and Saito, T. (2012). Programmed cell death 1 forms negative costimulatory microclusters that directly inhibit T cell receptor signaling by recruiting phosphatase SHP2. J Exp Med 209 , 1201-1217.

Yoon, S.O., Zhang, X., Berner, P., Blom, B., and Choi, Y.S. (2009). Notch ligands expressed by follicular dendritic cells protect germinal center B cells from apoptosis. J Immunol 183 , 352-358.

Yu, J., Angelin-Duclos, C., Greenwood, J., Liao, J., and Calame, K. (2000). Transcriptional repression by blimp-1 (PRDI-BF1) involves recruitment of histone deacetylase. Mol Cell Biol 20 , 2592-2603.

Zal, T., Volkmann, A., and Stockinger, B. (1994). Mechanisms of tolerance induction in major histocompatibility complex class II-restricted T cells specific for a blood-borne self-antigen. J Exp Med 180 , 2089-2099.

Zanoni, I., Spreafico, R., Bodio, C., Di Gioia, M., Cigni, C., Broggi, A., Gorletta, T., Caccia, M., Chirico, G., Sironi, L. , et al. (2013). IL-15 cis presentation is required for optimal NK cell activation in lipopolysaccharide-mediated inflammatory conditions. Cell reports 4, 1235-1249.

228

Zeng, R., Spolski, R., Finkelstein, S.E., Oh, S., Kovanen, P.E., Hinrichs, C.S., Pise- Masison, C.A., Radonovich, M.F., Brady, J.N., Restifo, N.P. , et al. (2005). Synergy of IL-21 and IL-15 in regulating CD8+ T cell expansion and function. J Exp Med 201 , 139- 148.

Zhang, L., Yu, Z., Muranski, P., Palmer, D.C., Restifo, N.P., Rosenberg, S.A., and Morgan, R.A. (2013a). Inhibition of TGF-beta signaling in genetically engineered tumor antigen-reactive T cells significantly enhances tumor treatment efficacy. Gene Ther 20 , 575-580.

Zhang, N., and Bevan, M.J. (2010). Dicer controls CD8+ T-cell activation, migration, and survival. Proc Natl Acad Sci U S A 107 , 21629-21634.

Zhang, T., Wu, M.R., and Sentman, C.L. (2012). An NKp30-based chimeric antigen receptor promotes T cell effector functions and antitumor efficacy in vivo. J Immunol 189 , 2290-2299.

Zhang, Z.N., Xu, J.J., Fu, Y.J., Liu, J., Jiang, Y.J., Cui, H.L., Zhao, B., Sun, H., He, Y.W., Li, Q.J., and Shang, H. (2013b). Transcriptomic analysis of peripheral blood mononuclear cells in rapid progressors in early HIV infection identifies a signature closely correlated with disease progression. Clin Chem 59 , 1175-1186.

Zhao, L., Ji, W., Zhang, L., Ou, G., Feng, Q., Zhou, Z., Lei, M., Yang, W., and Wang, L. (2010). Changes of circulating transforming growth factor-beta1 level during radiation therapy are correlated with the prognosis of locally advanced non-small cell lung cancer. J Thorac Oncol 5, 521-525.

Zhao L, J.W., Zhang L, Ou G, Feng Q, Zhou Z, Lei M, Yang W, Wang L (2010 ). Changes of circulating transforming growth factor-beta1 level during radiation therapy are correlated with the prognosis of locally advanced non-small cell lung cancer. J Thorac Oncol 5, 521-525.

Zhou, L., Ivanov, II, Spolski, R., Min, R., Shenderov, K., Egawa, T., Levy, D.E., Leonard, W.J., and Littman, D.R. (2007). IL-6 programs T(H)-17 cell differentiation by promoting sequential engagement of the IL-21 and IL-23 pathways. Nat Immunol 8, 967- 974.

229

Zhu Y, J.S., Chen E, Dai S, Li C, Morel P, Liu L, Zhang X, Lu B. (2010 ). T-bet and eomesodermin are required for T cell-mediated antitumor immune responses. J Immunol. 185 , 3174-3183.

Zou, W. (2005). Immunosuppressive networks in the tumour environment and their therapeutic relevance. Nat Rev Cancer 5, 263-274.

230

Biography

Regina Lin was born in Singapore on May 23, 1985. Regina went to college at the

National University of Singapore (NUS), where she majored in Life Sciences with a concentration in Biomedical Science. In 2008, Regina obtained her Bachelor of Science degree with 1 st Class Honors. After taking a year off school to work as a research assistant, Regina began her graduate education at Duke University’s Department of

Immunology in 2009, where she pursued her PhD in the laboratory of Dr. Qi-Jing Li.

Under Dr. Li’s mentorship, Regina published two peer-reviewed first-author papers

“Biological evaluation of subglutinol A as a novel immunosuppressive agent for inflammation intervention” in ACS Medicinal Chemistry Letters and “Targeting miR-23a in CD8+ cytotoxic T lymphocytes prevents tumor-dependent immunosuppression” in the

Journal of Clinical Investigation”, and published an adjunct author’s view “miR-23a blockade enhances adoptive T cell transfer therapy by preserving immune-competence in the tumor microenvironment” in OncoImmunoogy. In addition, Regina was also a contributing author to two other peer-reviewed manuscripts “Mucosal-resident T lymphocyte Repertoire Diversity Predicts Clinical Prognosis in Gastric Cancer Patients” and “Association of CD8+ T Lymphocyte Repertoire Spreading with the Severity of

DRESS Syndrome”. At Duke University, Regina was awarded the Chancellor’s

Scholarship in AY2009/2010 and came in first place at the Bernard Amos Poster

Competition in 2014. Upon completing her PhD degree at Duke Immunology, Regina hopes to embark on a career on the research and development of novel immunotherapeutic targets and approaches. 231